NHS Making Data Count and the NHS SPC tool

NHS England's programme for replacing RAG-rated board reports with control charts. What it teaches, how its conventions differ from industrial SPC, and the free tools that implement it.

NHS England · Our reading of it — the source document is the authority. The document itself ↗

Last reviewed 2026-08-18

Making Data Count is NHS England's programme for getting statistical process control into NHS board and operational reporting, in place of red-amber-green ratings and month-on-month comparison. It is the largest deliberate SPC adoption programme anywhere, in any sector: well over a hundred organisations had been through the training by 2021 and it has continued to spread since — including, in one evaluated trial, to control sites that were never trained, simply because the charts were persuasive.

The case it makes is not primarily statistical. It is that a board asked to explain every red is being asked to explain noise, and that the explanations generated are worse than useless because they attribute causes to variation that has none.

What it teaches

  • Two kinds of variation. Common cause (the system doing what the system does) and special cause (something changed). Reacting to the first is "tampering" and reliably makes performance worse.
  • The XmR chart for nearly everything. Healthcare data is mostly one number per week or month, so the individuals-and-moving-range chart is the workhorse. Mean, and process limits at three sigma estimated from the mean moving range.
  • Four rules, and no more. A point outside the process limits; a run of points on one side of the mean; a run of consecutive points rising or falling; and two of three points close to a limit. The team is explicit that adding more rules buys false positives, not sensitivity.
  • Icons instead of colours. A variation icon (is this common or special cause, and is the special cause improvement or concern?) and an assurance icon (will this system reliably hit the target?), reported side by side. Splitting those two questions is the programme's single best idea: "consistently missing the target" and "wildly variable" are different problems with different fixes.
  • Targets are not limits. The target is drawn as its own line. Process limits come from the data; targets come from a policy. Confusing them is the original sin of performance reporting.

The tools

What Notes
NHS England SPC tool The official Excel tool, free, with the icon logic built in. There is a multiple-chart version for a whole board pack.
Making Data Count guides The core training PDFs (Part 1 and Part 2), published by NHS England and widely mirrored.
NHS Learning Hub catalogue The training catalogue and sessions.
NHSRplotthedots The NHS-R community R package that draws Making Data Count XmR charts, with a tutorial from NHS England's analytics team.

If you are producing a monthly board pack, use these. They are free, they are the convention your board has been trained on, and no commercial tool will match them for that job.

Where the conventions differ from industrial SPC

Worth knowing before you buy anything built for manufacturing:

  • Run length. Making Data Count typically flags a shift at a run of six or seven points on one side of the mean; trusts vary. The industrial convention (Nelson rule 2) is nine. Same idea, different sensitivity — the NHS charts will flag a sustained shift sooner.
  • Icons versus rule numbers. An NHS chart reports a variation icon; an industrial chart reports "rule 2, points 14–22". The information is the same and the audience is not.
  • Chart family. Manufacturing uses subgrouped charts heavily (X̄-R, X̄-s) and a full attribute family (p, np, c, u). NHS practice is overwhelmingly XmR, with funnel plots for comparing organisations.

Where a monitoring tool fits alongside this

The NHS tools are analysis and reporting tools: an analyst opens a workbook, pastes data, and produces a pack. That is the right shape for a board report and the wrong shape for anything that needs watching between meetings.

What a hosted tool adds is the part Excel structurally cannot do: data that arrives on its own, limits frozen with their baseline recorded, rules evaluated as each point lands, an alert when one trips, and a screen a ward or department can leave running. See SPC for healthcare for a worked example and, importantly, for the list of things this platform does not do — including the NHS icons and the six-point run rule.

We would rather tell you that up front than have it surface in a board meeting.

Who this applies to

  • SPC for healthcare and the NHS — NHS England's Making Data Count replaced RAG ratings with control charts for a reason. What healthcare SPC looks like, where the NHS conventions differ from the industrial ones, and what this platform does and does not do about it.

Charts that hold still long enough to be audited

Limits computed from a baseline window and frozen there, every change to them recorded with who made it and when, the rule set stated on the chart, and a read-only view anyone can be sent. That is most of what these documents ask a tool to do.