Assess, plan, do, review is supposed to be a cycle. In too many schools, it survives as four headings copied into a support plan.
That difference matters. Headings describe a process. A reliable evidence trail shows that the process happened: what staff knew at the start, what they decided to try, what was delivered, what changed and what the next decision was.
The SEND Code of Practice describes SEN support as a graduated approach built around assess, plan, do and review. For schools, the practical task is to make each round traceable without turning every review into an archaeological dig through emails, old Word files and intervention spreadsheets.
In brief: a sound assess, plan, do, review record links each review to a clear baseline, names the agreed provision and responsible staff, records what was actually delivered, compares progress with the starting point and preserves the reason for the next decision.
Why four completed boxes are not an evidence trail
A plan can look finished while saying very little.
Consider this review note:
Some progress has been made. Continue support and review next term.
It is tidy, positive and almost useless. It does not tell the next teacher what progress means. It does not say whether the planned support happened, whether the pupil's needs changed or why the same approach should continue.
A stronger record might say:
At the September baseline, Sam needed an adult prompt for each stage of a three-part written task. The plan introduced a visual task strip and a five-minute pre-teach twice a week. By the November review, lesson records showed that Sam usually began the second and third stages independently when the task strip was used. Pre-teach took place in seven of the planned eight sessions. The team will keep the task strip, reduce adult prompting and test whether one weekly pre-teach session is enough.
That paragraph is not longer for the sake of it. Each sentence does a job. It connects the starting point, provision, delivery, outcome and next decision.
What should the assess stage record?
Assessment is not a list of labels. It is the best current account of the pupil's needs, strengths, barriers and response to support.
Start with evidence that can be dated and checked:
- pupil and parent views
- teacher assessment and work samples
- attendance or participation patterns where relevant
- observations recorded for a clear purpose
- previous support and the pupil's response
- advice from specialists or other professionals
The aim is not to collect everything. It is to establish the baseline for the decision you are about to make.
A useful baseline is specific enough to compare later. "Finds writing difficult" is too broad. "Writes one or two independent sentences before stopping and needs verbal prompting to return to the task" gives the review something to test.
Keep source and interpretation separate. A pupil's exact words should remain their words. A teacher's view should be attributed to the teacher. A professional report should not quietly become a school diagnosis. This makes the record more honest and easier to review.
What makes the plan stage useful to staff?
A plan should settle five questions:
- What outcome are we working towards?
- What provision or adjustment will be used?
- Who is responsible for it?
- How often, where and for how long should it happen?
- What evidence will the review use?
If one of those answers is missing, the plan is likely to drift.
Targets also need a visible connection to the baseline. "Improve reading confidence" sounds reasonable but cannot be measured consistently. A better target describes the behaviour or outcome staff expect to see, under what conditions and by when.
Avoid filling the plan with every sensible classroom strategy. Choose the support that responds to the assessed barrier. A shorter plan that teachers can use is better than a comprehensive document that nobody can carry into practice.
What belongs in the do stage?
The do stage is where many evidence trails break. Schools record the plan and the review, but little about what happened between them.
This does not require daily essays. It requires enough delivery evidence to distinguish three different situations:
- the support happened and appeared to help
- the support happened but did not produce the expected change
- the support was not delivered consistently enough to judge
Those conclusions lead to different decisions. Without a delivery record, they collapse into the same vague review.
Agree a proportionate method before the cycle starts. That might be a session count, a brief teacher check, a work sample at set points or a simple record of whether an agreed adjustment was available. Choose evidence that staff can maintain during the normal school week.
The class or subject teacher remains central. An intervention delivered elsewhere does not remove the need to connect learning back to the classroom. Record how the pupil responds in ordinary lessons, not only in the intervention room.
How should the review compare progress?
A review should compare like with like wherever possible.
Return to the original baseline and ask:
- What changed against the agreed outcome?
- What evidence supports that conclusion?
- Was the provision delivered as planned?
- What did the pupil and parent notice?
- Did any new need, strength or barrier emerge?
- What should stop, continue or change?
This is the point where version history matters. If the latest document overwrites the previous plan, the school loses the before-and-after view. Staff can see today's wording but not the decision it replaced.
MeritDocs supports review versions that can be linked to a previous baseline. That gives staff a clearer account of what the new review is building on, while leaving professional judgement with the SENCO and teaching team.
A practical APDR record structure
Use six linked entries for each cycle:
- Baseline: date the source evidence and describe the pupil's current position, including pupil and parent views.
- Decision: state the barrier, the chosen response and who agreed it.
- Delivery: name the provision, responsible staff, frequency, duration and classroom connection.
- Measure: decide in advance what evidence will show change.
- Record: note what happened and any material disruption to delivery.
- Outcome: compare with the baseline, explain the next decision and name its owner and review date.
This is compact enough to use but strong enough to explain the school's reasoning months later.
How version control changes the quality of review
Shared drives make it easy to save another file. They do not make it easy to know which file is current, what changed or who approved the change.
A dependable SEND workflow needs more than filenames such as `Support Plan FINAL v4 new.docx`. It needs an approved baseline, a new review version, a record of the change and one current document that staff can find.
MeritDocs brings document versions, review context and lifecycle state into one governed workflow. The practical gain is not a prettier archive. It is the ability to see what the review changed without reconstructing the pupil's history from disconnected files.
Schools should still apply sensible access controls. Not every member of staff needs every document. Staff do need the current information required to support the pupil safely and consistently.
Common APDR mistakes to fix
Do not rewrite the whole plan at every review because that hides what changed. Do not judge impact without checking whether the provision happened. Agree the review measure during planning, preserve who said what, and give every next action an owner and date. The current plan tells staff what to do now. The version trail explains why the decision was made.
What will inspectors and leaders want to understand?
Ofsted's school inspection operating guide identifies inclusion as a central area of evidence gathering. A school should not build APDR records as an inspection performance.
It should be able to explain its ordinary practice: how needs are identified, how provision is chosen, whether it is delivered, what difference it makes and how leaders know.
A small number of complete, typical cycles will usually tell that story better than a huge folder assembled the night before a visit. Good records follow the work. They do not try to replace it.
FAQ
How often should assess, plan, do, review happen?
The timing should reflect the pupil's needs, the provision and the period needed to judge impact. Set the review date when the plan is agreed rather than letting a generic termly timetable delay an urgent response.
Can AI write the review for the SENCO?
AI can help structure recorded evidence. It should not invent progress, infer unsupported needs or make the professional decision. Staff must check the sources and approve the record.
The takeaway
Assess, plan, do, review works when each cycle leaves the next team with a better starting point. The record should show what staff knew, what they tried, what happened and why they changed course.
MeritDocs is designed to keep that chain intact through structured SEND documents, baseline-linked review versions and a current record staff can find. The technology does not make the judgement. It stops the judgement disappearing inside another file.