Measurement-based care works in trials. Twenty-four years on, it still rarely lands

Measurement-based care is the practice of administering a standardised symptom questionnaire at each visit and using the score to guide the next treatment decision. Over the past two decades, research has shown that it improves clinical outcomes in mental health care — and translating it into routine practice remains challenging [s1].

A systematic review published in Psychiatric Services attempts to establish what actually works to get it adopted. Its most informative number is how little evidence it found.

Sixteen studies from two decades

The review searched MEDLINE, PubMed, Web of Science, Google Scholar and PsycNet for evidence gathered between 2000 and 2024 on interventions supporting measurement-based care implementation in adult mental health services [s1]. The searches yielded 1,120 records in a first pass covering 2000–2022 and 1,095 in a second covering 2023–2024 [s1].

Sixteen articles survived full-text assessment and were critically appraised using the Mixed Methods Appraisal Tool [s1]. Quantitative descriptive designs were the most common study type [s1].

Sixteen studies drawn from two searches returning 1,120 and 1,095 records [s1], across twenty-four years and an entire health sector, is a thin base on which to build implementation guidance — and "quantitative descriptive" means most of them observed rather than tested. That is the finding beneath the finding: the effectiveness question has been studied far more than the adoption question.

What the review reports

Where measurement-based care was implemented, the review links it to symptom reduction, improved medication adherence and increased use of self-report tools [s1]. Patient satisfaction rose when the practice supported shared decision making [s1].

Providers viewed it positively, but feedback practices varied and concerns about relational impact remained [s1] — the worry, familiar from clinical debate, that a questionnaire at the start of a session changes what the session is. The review also records a professional split: psychologists engaged more with measurement-based care than did psychiatrists [s1].

On what drives uptake, system-level strategies and electronic health record integration increased adoption, but broad adoption remained difficult [s1]. The barriers listed are time constraints, lack of training, hierarchical structures, negative attitudes and technical limitations [s1].

The authors conclude that effective implementation requires context-sensitive strategies, interprofessional collaboration, and active involvement of patients, service users and caregivers, and that future research should focus on underrepresented settings, diverse professional roles and hybrid designs, with standardised measures and detailed reporting [s1].

That conclusion is unobjectionable and also unspecific, which is what a review of sixteen heterogeneous descriptive studies can honestly support.

What it looks like in one clinic

A separate analysis published a few weeks earlier shows the other side: what the data look like when measurement-based care is running at scale in a single service.

Researchers analysed electronic health records of 4,778 patients aged 12 and over from an academic outpatient psychiatry clinic, requiring a baseline PHQ-9 score above 9 and at least one follow-up PHQ-9 measurement [s2]. Factors associated with response and remission were evaluated using Cox regression models adjusted for demographic and clinical covariates [s2].

Response and remission rates were 46.3% and 28.2% respectively [s2].

For response, the significant associated factors were combined treatment (HR 1.189, p=0.014), follow-up PHQ-9 administered within four to nine months (HR 1.303, p<0.001), and post-traumatic stress disorder status, reported in the paper as absence of PTSD with HR 0.817, p=0.018 [s2]. For faster remission the factors were combined psychiatry and psychotherapy treatment (HR 1.189, p=0.014), timely follow-up PHQ-9 administration within four to nine months (HR 1.255, p=0.002), lower baseline PHQ-9 scores (HR 0.931, p<0.001), and again PTSD status, reported as absence of PTSD with HR 0.751, p=0.010 [s2]. Alternative remission definitions — PHQ-9 below 12 or below 9 — yielded similar findings [s2].

Two of the reported hazard ratios are labelled as an absence of PTSD while carrying values below 1, a combination that is difficult to interpret from the summary alone; the consistent substantive claim in the paper is that comorbid PTSD matters to depression outcomes [s2].

What the clinic data can and cannot show

This is a retrospective analysis of records from one academic clinic, with an inclusion rule that required at least one follow-up PHQ-9 [s2]. That rule builds in a selection effect that runs in a specific direction: patients who disengaged after a single visit are, by construction, absent. A 46.3% response rate is a response rate among patients who came back.

The design also cannot separate the effect of measurement from the effect of treatment. That timely follow-up PHQ-9 administration within four to nine months associates with better outcomes [s2] is compatible with monitoring improving care, and equally compatible with patients who attend follow-up appointments doing better for reasons unrelated to the questionnaire. Nothing in a records-based cohort distinguishes those.

What the association with combined psychiatry and psychotherapy treatment [s2] adds is more modest than it first appears, since the patients who receive both differ systematically from those who receive one.

The gap the two papers frame

Read together, the papers describe a familiar shape in mental health services research. The practice has evidence behind it [s1]. A clinic that runs it generates data detailed enough to identify who responds and how fast [s2]. And after twenty-four years, the literature on how to get services to adopt it amounts to sixteen studies, mostly descriptive [s1].

The review's own recommendation — that future work use hybrid designs, which test an intervention and its implementation simultaneously — is the concrete part of its conclusion [s1]. It is also an admission that the field has been running those two questions separately, and that the answer to the first has not produced the second.

Sources

• Implementation of Measurement-Based Care in Mental Health Services: A Systematic Literature Review, Psychiatric Services, published online 8 July 2026
• Clinical Characteristics Associated With Response and Remission From Depression Using Measurement-Based Care, Psychiatric Research and Clinical Practice, published online 16 June 2026