Research desk · Running your practice
Measurement-based care sounds like something a clinic does with a dashboard and a data team. The research says the active ingredient is smaller than that: a short measure, given every session, that you actually look at with the client. Here is what the trials found, who benefits most, and a setup that fits a 50-minute hour.
Measurement-based care (MBC), sometimes called routine outcome monitoring or feedback-informed treatment, has three parts. The client completes a brief, validated measure at regular intervals, usually every session. You review the score, ideally against the client’s own trend. Then you talk about it together and use it to decide whether to keep going, adjust, or rethink. [6]
It is not an intake screener you never repeat. It is not a quarterly questionnaire that lands in a portal no one opens. A review in Psychiatric Services made the point directly: one-time screening, infrequent assessment, and feedback delivered outside the clinical encounter were the versions that did not work. [6] If the number never enters the room, it is paperwork.
The honest headline is that the average effect is small, and the effect where it matters most is meaningful.
A 2021 meta-analysis in Clinical Psychology Review pooled 58 studies with 21,699 patients. [1] Progress feedback improved symptom outcomes with a small effect overall (d = 0.15, 95% CI 0.10–0.20). For clients flagged as not on track, the effect was d = 0.17. Feedback was also linked to a modest increase in the odds of dropout (OR 1.19), which is worth knowing before you assume more measurement always means more engagement.
A 2018 meta-analysis by Lambert and colleagues covered 24 studies of routine outcome monitoring. Roughly two-thirds found an advantage for feedback. [2] In clients predicted to have a poor outcome, feedback reduced deterioration and nearly doubled the rate of reliable or clinically significant improvement.
The largest single trial in routine care is a cluster RCT from England’s NHS talking-therapies services, published in Lancet Psychiatry. [3] It involved 79 therapists and 2,233 patients. Among patients who were not on track, those whose therapists got feedback had better depression (PHQ-9, d = 0.23) and anxiety (GAD-7, d = 0.19) outcomes than those whose therapists did not.
The newest trial looks at speed rather than end points. In a 2025 RCT in JAMA Network Open, 154 adults with major depression in Pakistan received antidepressant treatment either guided by measurement or as usual. [4] The measurement-guided group reached response in a median of 2 weeks versus 4, and remission in 4 weeks versus 8. By week 24 the groups did not differ in response or remission. That is a medication study, not a psychotherapy one, but the pattern is familiar: measurement helps you notice sooner and act sooner. Weeks of suffering saved are real even when the finish line looks the same.
If you remember one thing from this literature, make it this: feedback earns its keep with the client who is quietly not getting better.
In-session impressions are filtered through a client who may want to please you or protect themselves. A score that moves the wrong way for three sessions running is a prompt you can’t easily talk yourself out of.
What do you do with that prompt? A 2026 study in the Journal of Consulting and Clinical Psychology analyzed 192 sessions from 44 not-on-track cases across 10 therapists who used PHQ-9 and GAD-7 feedback. [5] The obstacles therapists and clients named most often were the impact of trauma, resistance or ambivalence, social and interpersonal problems, stressful life events, and problems in the therapy process itself. The responses that came up most were practical: re-agree on goals and tasks, narrow the focus, build an individualized formulation, and work on expectations and motivation.
In other words, the score does not tell you what is wrong. It tells you to stop and ask.
For a general outpatient caseload, two free, short measures cover a lot of ground.
| Measure | What it tracks | Useful anchors |
|---|---|---|
| PHQ-9 | Depressive symptoms, 9 items, scored 0–27 | 5, 10, 15, and 20 mark mild, moderate, moderately severe, and severe. A score of 10 or more had 88% sensitivity and 88% specificity for major depression in the validation study. [8] A 5-point change is a reasonable threshold for meaningful individual change. [9] |
| GAD-7 | Generalized anxiety symptoms, 7 items, scored 0–21 | At the recommended cut point of 10, sensitivity was 89% and specificity 82% for GAD in the validation study. [10] |
The forms and scoring instructions are published at phqscreeners.com. If your work is mostly couples, trauma, substance use, or children, pick a validated measure built for that population instead of forcing these two to fit.
Two cautions. First, item 9 on the PHQ-9 asks about thoughts of death or self-harm. If you give it every session, you need a plan for reviewing it before the client leaves, every time. Second, cut points are for screening populations. For tracking one person, the trend and the size of change matter more than which band they are in.
For “did this client really change?” the classic tool is Jacobson and Truax’s reliable change index, which asks whether a change is larger than what the measure’s own error would produce. [11] You don’t need to compute it by hand; the practical version is to pick a change threshold for your measure in advance and hold yourself to it.
Opening the conversation does not have to be clinical-sounding:
“Your score went from 16 to 11 since we started. Does that match how the last few weeks have felt, or is the form missing something?”
And when things are stuck:
“These numbers haven’t moved much in a month. I don’t read that as you doing anything wrong. I read it as a sign we should check whether we’re working on the right thing.”
That second script matters. The dropout signal in the meta-analysis is a reminder that a flat or rising score can feel like a verdict to a client. [1] Frame it as information about the plan, not the person.
A 2019 review in JAMA Psychiatry estimated that fewer than 20% of behavioral health practitioners use measurement-based care routinely. [7] The barriers it described sit at every level: clients who find forms tedious, clinicians who worry measures will reduce therapy to a number, organizations without workflows, and systems that do not pay for it.
In solo practice you are the organization and the system, which cuts both ways. Nobody will build the workflow for you, but nobody needs to approve it either. Most practice-management systems can send forms before sessions; check what yours supports before buying anything new. Our practice systems guide covers how to keep tools like this from multiplying.
The worry that numbers will flatten the work is fair. The bigger risk, per the studies above, is continuing the same plan with a client who is not improving. The measure is a two-minute second opinion.
Each card links to the PubMed record. Numbers are reported the way the authors reported them. Read the abstract before you quote any of them.
de Jong K, Conijn JM, Gallagher RAV et al.
58 studies, 21,699 patients. Progress feedback improved outcomes overall (d = 0.15, 95% CI 0.10–0.20) and for not-on-track cases (d = 0.17). Dropout odds were modestly higher with feedback (OR 1.19).
Lambert MJ, Whipple JL, Kleinstäuber M
24 studies of routine outcome monitoring. About two-thirds favored feedback. In predicted poor-outcome cases, feedback reduced deterioration and nearly doubled reliable or clinically significant change.
Delgadillo J, de Jong K, Lucock M et al.
79 therapists, 2,233 patients in NHS services. Feedback improved outcomes for not-on-track patients: PHQ-9 d = 0.23, GAD-7 d = 0.19.
Husain MI, Nigah Z, Ansari SUH et al.
154 adults with major depression, antidepressant care. Measurement-based care halved median time to response (2 vs 4 weeks) and remission (4 vs 8 weeks); no difference in response or remission at week 24.
Williams E, Christensen H, Aguirre E et al.
10 therapists, 44 not-on-track cases, 192 sessions. Common obstacles: trauma, ambivalence, interpersonal problems, life stressors, therapy-process problems. Common fixes: re-agree goals and tasks, sharpen focus, individualize the formulation, work on expectations and motivation.
Fortney JC, Unützer J, Wrenn G et al.
The case for MBC as a standard. Notes that one-time screening, infrequent assessment, and feedback outside the clinical encounter were ineffective.
Lewis CC, Boyd M, Puspitasari A et al.
Implementing MBC in behavioral health. Fewer than 20% of practitioners use it routinely; barriers at patient, practitioner, organization, and system levels.
Kroenke K, Spitzer RL, Williams JB
PHQ-9. Scores of 5, 10, 15, and 20 represent mild, moderate, moderately severe, and severe depression. A score of 10 or more: 88% sensitivity and 88% specificity for major depression.
Löwe B, Unützer J, Callahan CM et al.
PHQ-9 for monitoring change. Supports a 5-point change as a clinically meaningful difference for individual patients.
Spitzer RL, Kroenke K, Williams JB et al.
GAD-7. At a cut point of 10, 89% sensitivity and 82% specificity for generalized anxiety disorder.
Jacobson NS, Truax P
Clinical significance and the reliable change index. A statistical way to judge whether an individual’s change exceeds measurement error.
Education for clinicians, not medical or legal advice. This page summarizes published research for licensed mental health professionals. It is not a treatment recommendation for any individual, and it does not replace your own clinical judgment, your licensing board’s rules, or consultation with a prescriber or attorney. Evidence reviewed through October 24, 2026. Spot an error? Tell us and we will fix it.
Keep reading
What predicts it, how it touches outcomes, and what helped.
ResearchRead →02What the 850-person trial found, and where the boundaries are.
ResearchRead →03What the newest trials and state laws mean for your practice.
ResearchRead →