School psychologists described a caseload governed by two clocks that don't agree: a school calendar that concentrates demand into particular months, and a queue of families waiting on documents. The write-up sits between them, and it's the part that determines how many assessments actually get finished.
I'm not a clinician and I've never worked in a school setting. What I did was ask several school psychologists to describe an ordinary term, and the same structure came back each time. This is their description, not my analysis of how the role should work.
Two clocks, and a queue of documents sitting between them.
What shapes the caseload?
Three pressures, applied at the same time.
The calendar. Referrals cluster, assessment windows cluster, and the meetings that depend on completed documents cluster. Work doesn't arrive evenly across a year, and the peaks don't move.
The queue. Families and teachers are waiting on something specific — a document that unlocks a decision, a placement, or a support. Every week of delay is a week of somebody waiting, and school psychologists feel that acutely in a way that a general throughput conversation doesn't capture.
Two audiences per document. The report is read by the family and by the school, and those readers want different things from the same pages. Writing for both is craft work, and craft work doesn't compress.
Layer those and you get a role where the assessment itself is the predictable part and everything around it is not.
Where is the bottleneck actually?
Not testing capacity, which is what an outsider would guess.
Every school psychologist I asked put it in the same place: the write-up. Testing gets scheduled and testing happens. The document waits for a block of uninterrupted time that a school day is not designed to produce. One person described the realistic options as before school, after school, or the weekend — and said the third one wins more often than they'd like.
That matters for anyone trying to fix it, because adding assessment capacity to a system whose constraint is the write-up doesn't increase output. It increases the queue. The same pattern shows up in private practice, and I've written about the arithmetic in what an assessment report actually costs a practice to produce.
What does the seasonal peak do to it?
It compounds, which is the part that surprised me.
A backlog isn't a static queue. A report written three weeks after the assessment costs more than one written the same week, because the case has to be reloaded from notes rather than recalled. So a peak doesn't just create a pile — it makes every item in the pile more expensive, at exactly the moment there's least room.
Several people described the same downstream consequence: a hard stop on intake past a certain point in the term. Not because assessment slots were full, but because the write-up queue was. That's a ceiling on how many children get seen, set by a documentation constraint rather than a clinical one.
A backlog doesn't just wait. It gets more expensive every week it waits.
What's different about the institutional buyer?
When a district or board is involved, the thing being bought is often a date.
Several people made a version of this point: for an institution, turnaround is the deliverable. Quality is assumed — it's the professional's job and nobody negotiates it — so the negotiable term becomes when the document arrives. Which puts the write-up, again, at the centre of a commitment somebody else has made.
It also changes what software procurement looks like. An institutional review asks a different set of questions than a solo clinician does, and the answers have to survive a process rather than a conversation.
What did they say actually helped?
Three things, none of which is working faster.
Protecting a writing block and treating it as unmovable. The most-cited change, and the hardest to hold, because a protected block has no person attached to it and everything else does.
Tracking the write-up separately from the assessment. Most people record assessment time as one figure. Splitting the write-up out is what makes the bottleneck visible in a staffing conversation rather than a private frustration.
Reducing what has to be produced from nothing. The structural scaffolding, the sections that repeat, the descriptive passages that carry no judgment — that's a real share of a long document, and it consumes a protected block before the thinking starts.
That last one is the only place I'd argue software belongs, and deliberately the only thing ours does: make sure the protected hour starts with something on the page. What the document concludes stays entirely with the person signing it.
If you're on the receiving end of one of these documents rather than writing them, who reads an assessment report and what each reader looks for covers the other side.
— Ian
