← All reportsAgent operating modelReport method

Agent-assisted analytical brief · Public aggregate data · RDY · Demonstration only

Worked example: AI-assisted MHSDS public-data briefing

MHSDS access and activity profile — six-month public-data brief (Nov 2025 to Apr 2026)

Public-data demonstration report: This is not an official Dorset HealthCare report. It uses public aggregate data only and requires human review and local owner confirmation before operational use.

What the agent was asked to do

Business question: From six months of public MHSDS Provider/RDY data, summarise access and activity measures, describe descriptive trends, and list what a mental health data owner must confirm.

Dataset: MHSDS Monthly Statistics — stacked Provider/RDY trend file

This is a first-draft analytical brief for human review — not an approved performance report.

Scope: Provider / RDY rows only — not Dorset population or ICB-resident views.

Data used

What this report can and cannot tell us

This report can:

This report cannot:

For that reason, this should be treated as a first-draft prompt for review, not a final performance judgement.

Headline reading

  • MHSDS access and activity profile for Dorset HealthCare (RDY as provider), Nov 2025 to Apr 2026.
  • Caseload-style measures (MHS23, MHS01) rose over six months while contact volume (MHS29) was broadly stable. This may indicate increased open-referral pressure, reduced contact intensity, coding effects, discharge effects, or case mix — local validation needed.
  • MHS69 shows volatile movement including a very large April increase — treat as a major validation flag, not operational improvement. Financial-year counting logic may apply.
  • MHS01 was renamed nationally from April 2026 to 'people with an open referral' (previously 'people in contact') — not necessarily proof of contact or access.
  • Provider and ICB-resident breakdowns must not be summed into one trust-wide headline.

Priority review flags

Caveats (read before interpreting)

Six-month headline measures (Provider/RDY)

Period: Nov 2025 – Apr 2026 (6 months, provisional MHSDS, Provider/RDY)

Descriptive statistics from trend_mhsds_access_rdy.csv; suppressed values excluded from averages and charts.

MeasureLatestPrevious monthMoM changeSix-month changeSix-month avgSix-month highSix-month lowTrend readingDesired direction / targetInterpretation
MHS23 — Open referrals at end of reporting period24,790 (Apr 2026)24,715 (Mar 2026)+75 (+0.3%)+875 (+3.7%)24,298.324,79023,765RisingNo simple targetRising open-referral stock; not automatically good or bad.
MHS01 — People with an open referral with services at end of reporting period21,860 (Apr 2026)21,755 (Mar 2026)+105 (+0.5%)+970 (+4.6%)21,319.221,86020,820RisingNo simple targetMay reflect caseload, activity, coding or discharge — not proof of access.
MHS29 — Contacts in reporting period31,305 (Apr 2026)31,240 (Mar 2026)+65 (+0.2%)+100 (+0.3%)30,490.831,37028,560Broadly stableNo simple targetActivity broadly flat or moving — not proof of access improvement.
MHS69 — CYP with at least two contacts (before 18th birthday)1,505 (Apr 2026)310 (Mar 2026)+1,195 (+385.5%)+1,170 (+349.3%)515.81,505260VolatileAccess-style metric; validate definitionMAJOR VALIDATION FLAG: very large month-on-month movement. MHS69 has financial-year counting logic — do not read as sudden operational improvement. Do not infer improvement without CYP data owner check — FY reset logic may apply.

Six-month trend charts

MHS23 — Open referrals at end of reporting period (Nov 2025 – Apr 2026)

Nov 25Dec 25Jan 26Feb 26Mar 26Apr 26

MHS01 — People with an open referral with services at end of reporting period (Nov 2025 – Apr 2026)

Nov 25Dec 25Jan 26Feb 26Mar 26Apr 26

MHS29 — Contacts in reporting period (Nov 2025 – Apr 2026)

Nov 25Dec 25Jan 26Feb 26Mar 26Apr 26

MHS69 — CYP with at least two contacts (before 18th birthday) (Nov 2025 – Apr 2026)

Nov 25Dec 25Jan 26Feb 26Mar 26Apr 26

Why this is useful

Human validation checklist

Bottom line

Over six months, open-referral stock measures rose while contact volume was broadly stable. That pattern may signal caseload pressure but cannot prove whether access is improving or worsening. MHS69 needs particular validation before any access finding. A mental health data owner should confirm definitions and whether these Provider/RDY figures match local reporting.

Audit trail and source checks

Six-month figures trace to stacked MHSDS main-data monthly files — no synthetic values.

Six-month statistics computed from trend_mhsds_access_rdy.csv (value_status=numeric rows only for averages/charts).

Key findings by review area (detail)

Key findings by review area

Stock measures (caseload-style)

MHS23 — Open referrals

Latest: 24,790. Six-month reading: rising. Open referrals are not new demand — higher counts are not automatically worse.

Human check: MHSDS/data owner to confirm open-referral definition.

MHS01 — People in contact

Latest: 21,860. Six-month reading: rising. Movement may reflect activity, discharge or coding.

Human check: Confirm local in-contact definition matches MHSDS.

Activity measures

MHS29 — Contacts in reporting period

Latest: 31,305. Six-month reading: broadly stable. Contact volume does not automatically mean better access.

Human check: Service/BI owner to confirm contact counting rules.

MHS69 — CYP with two contacts

Latest: 1,505. Confirm cohort and resident vs provider scope.

Human check: CYP mental health lead.

Draft interpretation (detail)

Draft interpretation

  • Six-month MHSDS brief (Nov 2025 – Apr 2026 (6 months, provisional MHSDS, Provider/RDY)) for RDY Provider rows.
  • Rising over the window: MHS23, MHS01.
  • Broadly stable: MHS29.
  • Descriptive trends only — not causal. Local MHSDS owner validation required before operational use.
Prompt excerpt
Load trend_mhsds_access_rdy.csv for MHS23, MHS01, MHS29, MHS69.
Compute six-month descriptive stats and trend readings.
No causal claims. Do not sum provider and resident breakdowns.
Technical audit — latest-month demo extract and source files

Latest demo month: 01/04/2026 to 30/04/2026. Six-month headline table uses stacked trend file only.

MEASURE_IDMEASURE_NAMEMEASURE_VALUEBREAKDOWN
MHS23Open referrals at the end of the reporting period24790Provider
MHS01People with an open referral with services at the end of the reporting period21860Provider
MH01People with an open referral with mental health services at the end of the reporting period19755Provider
AMH23Open referrals (adult mental health services) at end of the reporting period17085Provider
AMH01People with an open referral with adult mental health services at the end of the reporting period14830Provider
MH01bPeople with an open referral with mental health services aged 19 to 64 at the end of the reporting period10490Provider
CYP23Open referrals (children's and young people's mental health services) at end of the reporting period5375Provider
MH01aPeople with an open referral with mental health services aged 0 to 18 at the end of the reporting period5175Provider
CYP01People with an open referral with children and young people's mental health services at the end of the reporting period4965Provider
MH01cPeople with an open referral with mental health services aged 65 and over at the end of the reporting period4090Provider

Time-series fallback file: rdy_mhsds_monthly_mhsds_monthly_performance-april-2026_time_series_MHSDS Time_Series_data_Apr_2016_MarPrf_2023.csv

Pipeline: site/public-data/05_download_historic_public_data.R; render: site/R/03_render_public_reports.R

Human review required: Confirm definitions, publication status and local owner sign-off before any operational use.

← Back to draft reports