Routine outcome monitoring is one of the most evidence-based ways to improve therapy outcomes — and one of the most inconsistently applied. This article looks at how UK mental health services use PHQ-9, GAD-7 and CORE-10 in practice, the common pitfalls that reduce completion rates to token levels, and how a well-designed practice management system removes the friction that stops measures being completed.
Why routine outcome monitoring matters
Meta-analyses consistently show that feeding progress data back to therapists and clients improves outcomes, especially for clients who are not on track. In practical terms, that means clients who might otherwise drift for four or five sessions before their therapist notices reliable deterioration are identified inside one or two sessions and their treatment plan is adjusted. In services that fund care against contracted outcomes, this is also the difference between a green report and a difficult conversation with a commissioner.
Yet in many services, measures are completed at intake, forgotten mid-way, and only revisited at discharge — too late to change the course of care and too late to catch problems that a mid-treatment score would have surfaced.
Choosing the right measure for your service
- PHQ-9 — a nine-item screener for depression severity, sensitive to change and widely accepted by NHS commissioners. Cut-offs at 5, 10, 15 and 20 map neatly to mild, moderate, moderately severe and severe.
- GAD-7 — a seven-item screener for generalised anxiety, used alongside PHQ-9 as the NHS Talking Therapies minimum dataset. Cut-offs at 5, 10 and 15.
- CORE-10 — a broader ten-item measure of psychological distress used widely in counselling and third-sector services. Especially useful where PHQ-9 / GAD-7 feel too diagnostic for the service model.
- WEMWBS and SWEMWBS — positive mental wellbeing scales, useful for prevention, coaching and social-prescribing contexts.
Most services will use a small combination. The mistake to avoid is asking clients to complete six measures at every session — response burden collapses completion rates and, ironically, the quality of the data you do collect.
Getting measures completed — every session
The single biggest lever on completion rates is automation. If measures are sent by the system before each session, prefilled with client details, and land in the therapist’s dashboard with a trend line, completion rates rise from around 40–50% to well above 85% in the services we work with.
Practical tips that consistently work:
- Send the invitation 24 hours before the session, not on the morning of. Clients complete measures better when they have a quiet moment, not when they are already running late.
- Offer a single tap-through link — no login, no download, no PDF.
- Make it very obvious that a therapist will read the answer before the session. This alone significantly increases completion.
- For long courses of treatment, use a small fixed set of measures per session rather than rotating through many.
Feeding scores back into the therapy room
Scores are only clinically useful if the therapist sees them before the session. Look for a system that shows the latest score alongside a session-by-session trend, flags reliable deterioration, and lets clinicians add a short reflective note without leaving the client record.
Where a client has been offered a measure and not completed it, that should also be visible so the therapist can open the conversation gently at the start of the session. Non-completion is itself a piece of clinical data.
Reporting to commissioners
Reliable change, recovery and improvement rates — calculated correctly against the appropriate clinical cut-offs — are the language commissioners speak. A modern platform should produce these at the click of a button, filtered by contract, service or therapist, and should show the underlying numerator and denominator so anyone reading the report understands what is being counted.
Beware headline rates without denominators. A 90% recovery rate calculated on 10% of clients who happened to complete measures at both ends is not the same story as a 60% rate calculated on 85% of the caseload — and commissioners increasingly know the difference.
Common pitfalls
- Measures used as gatekeeping. If clients understand a score to be the reason they were discharged, completion collapses. Use measures clinically, not administratively.
- Rotating measures. Comparing session 3 CORE-10 to session 8 PHQ-9 is not clinically meaningful. Pick a set and stick to it for the episode.
- Manual scoring. If a therapist has to add scores by hand, errors and shortcuts creep in. Insist on automated scoring.
- Score without narrative. Numbers alone do not tell the story. The best platforms make it easy to attach a one-sentence clinical reflection to each score.
How Rhadar helps
Rhadar builds PHQ-9, GAD-7, CORE-10 and configurable service-specific measures directly into the client journey. Invitations, scoring, trend visualisation and commissioner-ready reporting are all included as standard on every plan, with no per-measure or per-message fees.
If you would like to see how it looks against a real client record, our team can walk you through it in about half an hour.
Working with clients who resist measures
A minority of clients push back on completing measures, particularly if they associate them with earlier gatekeeping experiences. The most effective response is almost always framing rather than enforcement: explain what the score will be used for, who will see it, and how it will shape the next session. Where a client remains uncomfortable, record that clinically and continue — a completed session with no measure is still better data than an abandoned episode.
For neurodivergent clients, or those with literacy or language barriers, offer a therapist-completed version at the start of the session as a conversation prompt. The measure then does double duty as a clinical tool and a rapport-building one.
Supervision and reflective practice
Aggregated outcome data is one of the most powerful reflective-practice tools available to a supervisor. Trends across a therapist’s caseload — particularly around drop-off, reliable deterioration and non-completion — open genuinely useful conversations that a case-by-case review cannot. Modern platforms should allow a supervisor to view a therapist’s aggregated (and appropriately anonymised) outcome data as part of routine supervision.
What good looks like
A mature service should expect: outcome-measure completion above 85% at session 1 and above 70% at every subsequent session; less than 5% of episodes with only intake and discharge measures; therapist-facing trend charts on every active client; and commissioner-ready recovery and reliable-improvement rates produced in under five minutes per contract. Getting from typical (40–50% completion, spreadsheets at month-end) to that standard is almost entirely a matter of workflow design and the right platform.