The NHS Matrix: where Æther came from
From 2015, three Matrix tools gave NHS mental health services in the south of England a current view of their own practice. Four of the views they offered are redrawn here with made-up data.
Æther began life as an internal project called SOMAT, short for "Son of Matrix". The Matrix was a family of assessment tools built for NHS mental health services in the south of England from 2015, and much of what Æther does was first worked out there. The screenshots we have from that time show real services' figures, so the four views below have been redrawn with made-up data.
The three Matrix tools
| Tool | For | Commissioned by | Recognition |
|---|---|---|---|
| EIP Matrix 2015 | Early intervention in psychosis teams in Time4Recovery, the south of England's EIP programme, assessed against NICE guidance | Oxford Academic Health Science Network | Its reports were used in parliamentary discussion by Norman Lamb MP, and it was central to the programme team's nomination for a BMJ Award in 2017 |
| PMH Matrix 2017 | Perinatal mental health across maternity, health visiting, talking therapies, specialist perinatal and adult community mental health services, recording how the NICE quality standard was being put into practice | Thames Valley Strategic Clinical Network | Endorsed by NICE, and a finalist and runner-up in the NICE Shared Learning Awards 2019 |
| CYP Matrix 2018 | Children and young people's mental health services in Surrey, Sussex and Kent, run by the NHS, county councils and charities, looking at caseload, workforce, skills, capacity and confidence | A group of clinical commissioning groups, led by Guildford and Waverley CCG | Designed with commissioners, clinicians and managers from the services that used it |
Before the Matrix, this information came from data that was collected, put together and written up into a report, often published months later, when it was too late to act on or to inform decisions about funding. The Matrix let services and commissioners see it while they could still do something about it.
Few questions, all relevant
The Matrix was deliberately light for the clinicians and managers who filled it in. There were few questions, and each one was something the person answering knew or could look up easily from their own service's records.
In the perinatal Matrix, each service was asked only about the parts of the NICE quality standard that applied to it. A maternity service answered questions about booking appointments and antenatal contacts, and a talking therapies service answered questions about how quickly treatment started. Neither saw the other's questions. Sessions stayed short, so they were finished. Because nobody was guessing at questions about someone else's service, the answers were accurate, and there were no sections left blank because they didn't apply.
| Quality statement | Maternity | Health visiting | Mental health | Talking therapies (IAPT) |
|---|---|---|---|---|
| 1. Valproate not prescribed to women who could become pregnant | Not asked | Not asked | Asked | Not asked |
| 2. Information about mental health in pregnancy at annual review | Not asked | Not asked | Asked | Not asked |
| 3. Asked about mental health at the booking appointment | Asked | Not asked | Not asked | Not asked |
| 4. Asked about emotional wellbeing at each routine contact | Asked | Asked | Not asked | Not asked |
| 5. Assessed when a mental health problem is suspected | Asked | Asked | Not asked | Not asked |
| 6. Psychological treatment starts within six weeks of referral | Not asked | Not asked | Not asked | Asked |
Results before review
At the end of a session, the team saw its own results straight away, each rated against national guidance, before sending them on for review. The team didn't need to work out what good looked like, because the rating told them, and the programme could mark an area it wanted every team to concentrate on that year.
| Measure | People | Rate | Rating |
|---|---|---|---|
| Physical health and support | |||
| Physical health check completed | 41 of 52 | 79% | |
| Smokers offered help to stop | 18 of 18 | 100% | |
| Families offered family intervention | 30 of 52 | 58% | |
| Carers offered an education programme | 33 of 40 | 83% | |
| Outcomes | |||
| Admitted to hospital | 9 of 52 | 17% | |
| Admissions under the Mental Health Act | 6 of 9 | 67% | |
| Readmitted within a year | 1 of 9 | 11% | |
Policy and practice
For each NICE quality standard, the perinatal Matrix asked two separate questions. Does the service have a policy in place, and, in a random sample of ten records from the quarter, did the thing actually happen? The answers often differed. Some areas had a policy and little sign of it in the records, and others were doing it without a written policy.
If a service couldn't answer, it was asked why, so the gaps were recorded too. A record held by another organisation or a system with no field for the answer is worth knowing about, and before this it would simply have been missing from the report.
Reasons given where records couldn't be checked: Area D, "notes held on paper at another site"; Area E, "our record system has no field for this".
Measured against the standard
Most charts showed the recommended level alongside the actual one, so the question was how far a service was from where it should be, and not just how it compared with its neighbours. The perinatal workforce charts showed each role's whole-time equivalents in post each quarter against the level recommended for that area's number of births, and a team losing staff showed up within a quarter or two instead of in next year's report.
Consultant psychiatrist
Community psychiatric nurse
Nursery nurse
One view across organisations
Perinatal mental health care in one area can involve a hospital trust's maternity service, a community trust's health visitors, a separate talking therapies provider and a specialist team, each covering a different patch. The perinatal Matrix brought them into one structure, so a clinical network could compare regions and then click through to the providers in a region and the services within each provider.
What Æther keeps
- Results go back to the person who gave them, then to review. In Æther, an evaluator reviews each session and can record their own answer beside the original, so both are kept.
- Each person is asked only what they can answer. Short, relevant sessions get finished, and the answers can be relied on.
- Results are measured against an agreed standard, and not only against each other.
- Gaps in the records count as findings. An answer that can't be given is worth knowing about.
- The organisation's structure decides what each person sees. Permission for a region covers the providers and teams within it, and you can click from a chart through to what's behind it.
- Rates take account of size, whether per 10,000 births or per flight observed, so large and small services can be compared fairly.
The difference in Æther is that the questions, the wording and the structure are no longer built into the software for one service. They're set for each organisation, which is how the same approach works on a flight deck now and could work on a customer call or a housing inspection.