Instruments, and how to read them
Scored instruments you can complete and interpret, alongside the ones you should know about but which are not reproduced here. Every entry carries its psychometrics, its licence, and the ways it is misused.
Why measure at all
Measurement in psychotherapy is not bureaucracy imported from medicine. It exists because clinical judgement about progress is unreliable in a specific and well-documented way: therapists substantially overestimate their own effectiveness, and identify only a minority of the patients who are actually deteriorating.
The case for routine outcome monitoring
- Between 5% and 10% of adults deteriorate during psychotherapy. Without measurement, most of those cases are missed.
- Therapists asked to rate their own outcomes place themselves, on average, far above the median — an effect that does not diminish with experience.
- Feeding progress data back to clinicians improves outcomes for not-on-track cases and reduces dropout, with effects strongest when a protocol exists for responding to an alert.
- Measurement also protects patients from treatments that are merely pleasant: a therapy both parties enjoy, that changes nothing, is very hard to detect from inside it.
Reading a score properly
| Concept | What it means | Why it matters |
|---|---|---|
| Reliable change index | Change larger than the measurement error of the instrument | Distinguishes real movement from noise |
| Clinically significant change | Crossing from the clinical into the functional population range | A large improvement that leaves someone still symptomatic is not recovery |
| Sensitivity / specificity | True positive and true negative rates at a cut-point | Both change with the population; a cut-point is not a property of the instrument alone |
| Positive predictive value | Probability that a positive screen is a true case | Collapses when the base rate is low — the central problem in suicide screening |
| Measurement invariance | Whether the instrument measures the same construct across groups | Without it, comparing scores across cultures or languages is not meaningful |
Depression
Anxiety
Trauma
Substance use
Process
General outcome
OQ-45
ReferenceOutcome Questionnaire-45
Track global distress across symptom, interpersonal and social role domains, session by session, with empirically derived alarm signals for cases that are going badly.
CORE-OM
ReferenceClinical Outcomes in Routine Evaluation — Outcome Measure
Thirty-four items across wellbeing, problems, functioning and risk; the standard outcome measure in UK psychological services.