The provider schedule template: slots, blocks, holds and release rules
In short
A provider schedule template is the rule set that turns a clinician's working week into bookable capacity: slots that may be filled, blocks that may not, holds reserved for a defined purpose, and release rules that decide when a hold becomes an ordinary slot. Reading a template as a grid of empty boxes is what produces both double-bookings and half-empty afternoons.
Key takeaways
- A template has 4 object types: slot, block, hold and release rule. Software that models only the first 2 cannot represent a real clinic day.
- A hold without a release rule is a permanent hole in the calendar, and it is the most common source of capacity that looks used but never was.
- Release should cascade — recall and waitlist demand first, then general booking — because unreleased urgent capacity is a wasted 30 minutes at 17:00.
- Overrides are not the enemy; unrecorded overrides are. Every slot created outside the template should carry a reason and an owner.
A template is not a picture of a week. It is a rule set answering, for any moment in a clinician's day: may something be booked here, of what kind, by whom, and from when. Four object types carry those answers — slot, block, hold and release rule — and software modelling fewer than 4 misrepresents every clinic that protects capacity.
The consequence shows up in booking behaviour. Read as a grid, protected capacity looks like free space, so the calendar gives away the urgent slots by 09:30. Read too conservatively, everything held looks unavailable, so patients are told there is nothing for 3 weeks while a third of the afternoon sits reserved for demand that never arrived.
Slot, block, hold, release rule
- Slot. A bookable interval with a visit type or a set of permitted types, a duration and an owning provider. It is the only object a patient ever occupies.
- Block. An interval that exists and cannot be booked: teaching, admin, theatre, travel between sites, lunch. A block is not absence — the provider is working, just not seeing booked patients.
- Hold. A slot reserved for a defined purpose and audience: same-day urgent, a specific referral source, a follow-up class, a clinic trial. A hold is bookable, but only by someone who satisfies the condition.
- Release rule. What converts a hold into an ordinary slot at a defined moment — 17:00 the previous working day, 2 hours before the slot, or when a named queue empties. Without one, a hold is a permanent hole.
Every hold is a bet that demand will arrive. The release rule is the point at which you stop paying for a bet that did not.
One clinic day written out as rules
| Time | Object | Rule attached |
|---|---|---|
| 08:30-09:00 | Block (admin) | Never bookable; excluded from utilisation denominators |
| 09:00-12:00 | 12 slots of 15 minutes | Follow-up visit types only; bookable up to 60 days ahead |
| 11:00 and 11:30 | 2 holds (same-day urgent) | Bookable only by triage until 17:00 the previous day, then released |
| 13:00-14:00 | Block (teaching) | Recurring weekly; overridable only by the clinical lead, with a reason |
| 14:00-17:00 | 4 slots of 40 minutes | New-patient types; requires completed intake, so minimum lead time of 2 days |
Written this way the day is machine-readable, and 3 questions that usually need a human become lookups: what can be offered to this patient, what is genuinely free, and what will free up if nobody claims it. The visit-type constraint in each row comes from a separate object, described in visit type as the unit a clinic day is built from.
Release rules are where the empty afternoons come from
Holds are the part clinics get wrong, almost always in the same direction: generous protection with no expiry. A fixed release hour is a start; the better pattern is a cascade that offers released capacity to queues already waiting before it becomes generally bookable.
- At release, offer to clinically due patients first. Those are recall entries, which behave differently from a cancellation waitlist — the difference is set out in recall, waitlist and hold as 3 queues.
- Then to the waitlist for that provider and visit type, in the order that queue defines rather than by who answers a phone first.
- Then to general booking, including any patient-facing channel that is permitted to see this visit type.
- Record what happened at each step. A hold that reaches general booking unclaimed every week is telling you it is too large, and that is only visible if release is logged.
Clinics often reach for overbooking instead, on the theory that some patients will not arrive. That is a different lever with a different failure mode, and before pulling it, confirm the attendance problem is real rather than a messaging one — the path is in reminders go out and the no-show rate does not move.
What any booking automation has to understand first
An agent offering appointments needs all 4 objects, not just the free intervals. Without holds it gives urgent capacity to a routine review; without release rules it calls a clinic full when 6 slots are about to open; without blocks it offers teaching time and gets quietly corrected afterwards. That correction is where the trust goes.
This bites hardest after hours, when nobody is there to correct anything, which is why template fidelity rather than speech quality usually decides answering service versus a voice agent that can book.
Most scheduling software models slots and blocks well and holds badly, so release logic ends up as a spreadsheet plus a habit. Building the missing piece is contained internal tools and operations work, sitting inside front desk, intake and patient access in a healthcare operations programme.
Frequently asked questions
Short answers to the follow-ups this page tends to raise.
What is a provider schedule template?
It is the rule set that turns a clinician's working pattern into bookable capacity. It is built from 4 objects: slots that may be filled, blocks that may not, holds reserved for a defined purpose and audience, and release rules that convert holds into ordinary slots. Templates usually recur weekly, with dated exceptions for leave and cover.
What is the difference between a block and a hold?
A block cannot be booked by anyone; a hold can be booked, but only by someone meeting a condition. Teaching time is a block. Two same-day urgent appointments kept back until the previous evening are holds. Holds need release rules and blocks do not, and treating a hold as a block is how clinics protect capacity they never use.
When should urgent-care holds be released?
Late enough to serve their purpose and early enough that someone can still fill them — commonly the end of the previous working day, with a second release a couple of hours before the slot. Release should cascade to recall and waitlist queues before general booking, and each release should be logged so the size of the hold can be reviewed against its actual use.
Who should be allowed to override a template?
A named, small group, with a reason recorded on every override. Clinics that forbid overrides get workarounds — appointments recorded in the wrong slot type, or in the notes — and clinics that allow silent ones stop being able to trust the template. The workable middle is override for a defined role, with a mandatory reason and a monthly review.
- scheduling
- capacity
- clinic operations
- patient access
The work behind this page
Builds from our portfolio that this page draws on.
Read next
- Visit type: the one field that decides duration, room, provider and prepVisit type is not a label on a calendar entry. It is the key every scheduling rule joins to, and where it is unmodelled no booking automation has anything to reason with.definition
- Recall, waitlist and hold: three different queues everyone calls one thingRecall fires on a clinical due date, waitlist fires on a cancellation, hold waits for the patient. Merge them and every automated message is wrong for two of the three.definition
- Double-booked by the booking agent: where the slot lock went missingTwo channels writing into one slot is not a scheduling bug. It is a missing reservation step between offering a time and committing it, and it is reproducible on demand.diagnostic
- Eligibility returns 'patient not found' for a clearly covered patientNot found is a statement about matching, not about coverage. Re-run the lookup with fewer identifiers, find which field broke the match, and stop retyping the member number.diagnostic
- Referrals come in by fax, email and portal, and none of them get bookedReferrals do not get lost because staff are careless. They get lost because a referral has no state, and something with no state has no age and no owner.diagnostic
- Reminders are sending and no-shows are flat: finding the step that failsA reminder can only fix forgetting. Join send logs to attendance, test 4 gates in order, and the flat number usually turns out to be a problem no wording change reaches.diagnostic
Working on something in this space?
Tell us where you are in a sentence or two. We'll tell you honestly whether we're the right team, and what a sensible first slice of the work looks like.
Start the conversation