# ND waits research: notes and caveats This file records how `nd_waits_research.json` was assembled, what could not be obtained, and which figures were deliberately left out. It is published alongside the dashboard because the gaps in the record are part of what the dashboard is about. Last revised 30 July 2026. ## Coverage summary All 14 territorial boards have at least a `pathway_status` entry. Nine boards have numeric figures for the assessment queue itself, meaning a list size, a referral or assessment count, or a wait in weeks: Ayrshire and Arran, Borders, Fife, Greater Glasgow and Clyde, Highland, Lothian, Orkney, Tayside, Western Isles. Eight boards have numeric figures for the separate ADHD medication queue that begins after diagnosis, all of them from the SPICe briefing: Ayrshire and Arran, Borders, Forth Valley, Grampian, Highland, Lothian, Orkney, Tayside. Three boards have no numeric figure of any kind and appear only through their pathway status: Dumfries and Galloway, Lanarkshire, Shetland. Only NHS Lothian publishes a list size and an annual assessment count for the same cohort, which is what a clearance rate requires. That is why the dashboard computes one for a single board rather than for fourteen. ## The SPICe briefing, and why reading it at first hand mattered **Resolved.** Earlier revisions of this file recorded that the briefing could not be retrieved and that everything attributed to it was second-hand. That is no longer true. The briefing has been obtained, read in full, and is committed at `site/sources/spice-sb-25-25.pdf`. The original obstacle is worth keeping on record, because it is the reason the dataset was wrong for a while. Every automated route returned a bot challenge rather than the document: - `digitalpublications.parliament.scot` direct URL: Cloudflare Turnstile check, both through the fetch tool and through `curl` with a current Chrome UA. - `www.parliament.scot` redirect target: the same Turnstile page. - `webarchive.nrscotland.gov.uk` mirror: the same challenge layer. - `web.archive.org`: blocked at the tool level, and 498/404 through `curl`. In the interim the briefing was quoted second-hand through an Aberdeenshire integration joint board report of 19 January 2026, which paraphrases it. Those second-hand records have since been deleted and replaced by first-hand ones by `etl/add_spice.py`. Reading the source directly changed three of fourteen board positions, all of them in the direction of the paraphrase being harsher than the briefing: - **NHS Grampian.** Recorded as having no adult pathway. The briefing says the Adult Autism Assessment Team continues in Moray and Aberdeen City, and that only Aberdeenshire withdrew. - **NHS Orkney.** Recorded as having no adult pathway. The briefing describes assessments bought in from an external provider while a pathway is developed. - **NHS Shetland.** Recorded as having no adult pathway. The briefing describes an adult autism assessment pathway that exists but is paused for funding. The distinction the briefing actually draws is narrower than "has a pathway": it is whether an adult who does **not** meet the criteria for referral to secondary mental health services can be assessed at all. Six boards are open on that basis, two of them for autism only. ### What was not taken from the briefing Its per-board figures appear as unlabelled bars in charts, with no data table. Nothing has been read off a bar. Only figures the briefing states in words or prints in its two tables are recorded here. As a check on that decision, and not as a source, the adult chart was measured programmatically against its gridlines. The bars summed to 23,128 against a stated national total of 23,339, within 1%. That confirms a reading would have been roughly right. It does not make it a citable number, and no measured value was added to the dataset. ## Sources that blocked automated access entirely - **whatdotheyknow.com** returns HTTP 403 behind a Cloudflare JavaScript challenge regardless of User-Agent, so no header spoofing through `curl` gets past it. This is still unresolved. It prevented searching WDTK for FOI responses from Ayrshire and Arran, Lanarkshire, Forth Valley, Greater Glasgow and Clyde, Shetland and Western Isles. Four of those six have since acquired figures from other sources: Ayrshire and Arran, Greater Glasgow and Clyde and Western Isles through the FOI tracker and board papers, and Forth Valley through the SPICe medication tables. **Lanarkshire and Shetland remain the two boards with no numeric figure from any source.** - **Scottish Parliament written Q&A search.** The search endpoint rendered a results page but the search term did not filter it, returning unrelated questions. No written answers with board-by-board figures were found. - **PHS CAMHS waiting times data quality document.** The content is solid and was read from a cached copy, but the exact publication URL could not be pinned down, so several `pathway_status` rows carry `https://publichealthscotland.scot/publications/` as a best-effort citation rather than a document link. - **NHS Orkney IJB papers, NHS Highland committee papers, NHS Borders FOI disclosure log, Aberdeenshire IJB papers.** The individual document URLs could not be confirmed live, so `source_url` for records drawn from them points at the organisation's general site rather than the specific PDF. The quoted content is verbatim from the documents; only the link is approximate. ## File mislabelling check `tayside_26067.txt` and `tayside_camhs.txt` were flagged as possibly mislabelled NHS Fife documents, because Carole Wilkinson's and Nicky Connor's names appear on them. **False alarm. Both are genuinely NHS Tayside documents.** - Both open with "NHS Tayside has now considered your request" and use the `IGTFOISA260xx` reference format, which is Tayside's prefix. - Both give the headquarters address as Ninewells Hospital and Medical School, Dundee, which is NHS Tayside's, not Fife's Victoria Hospital in Kirkcaldy. - NHS Tayside's own FOI disclosure log lists both by reference number with download links on `nhstaysidecdn.scot.nhs.uk`. - Nicky Connor is NHS Tayside's Chief Executive as of these documents, having moved from the same post at NHS Fife. That is the likely source of the original suspicion. ## Figures considered unreliable, or deliberately excluded - **NHS Borders FOI 622-18 date.** The response carries no explicit date. The reference format is *inferred* to mean 2018 by analogy with Tayside's convention, which is not stated in the document. Recorded as year only, and the record carries `date_inferred: true`. Treat it as old: it probably predates the current neurodevelopmental pathway framing entirely. - **NHS Highland's "> 1800".** The source states a lower bound, not a count. The record carries `qualifier: "minimum"` and the dashboard renders it as 1,800+. - **NHS Tayside's 177 and 210 week figures.** These are explicitly 90th percentile waits, stored as `percentile90_wait_weeks` rather than as a maximum or an average, so they cannot be silently compared with other boards' means. - **Aberdeenshire "approximately forty-five weeks".** From `aberdeenshire_mar2025.txt` paragraph 4.8.4, but in the context of general adult psychology and psychotherapy staffing rather than clearly the neurodevelopmental pathway. Excluded as ambiguous. - **Greater Glasgow and Clyde referral totals (8,906 / 8,477 / 6,994 / 7,100 for 2022 to 2025).** These are whole-of-CAMHS referral counts, not neurodevelopmental ones, even though part of the fall is attributed to neurodevelopmental patients moving off CAMHS books. Recording them as `referrals_12m` would misrepresent a general figure as a specific one. - **Borders "43% of total waiting list is external".** A genuine Borders figure, but a percentage of an unstated base, and it is unclear whether the external waiting list means the neurodevelopmental one or externally commissioned CAMHS activity generally. Excluded rather than forced into a measure it does not fit. ## Priority gaps 1. **Lanarkshire and Shetland**, the two boards with no numeric figure from any source. 2. **WhatDoTheyKnow**, which needs a real browser session rather than an automated fetch. 3. **A second assessment count.** Only Lothian publishes both a list size and an annual assessment rate. One more board would turn the clearance rate from a single observation into a comparison.