From the Editor
Should mental health clinicians embrace measurement-based care? Or is mental health becoming too technical (and forgetting patients as a result)?
When we speak about the future of mental health, we often think in terms of biomarkers and genetically-tailored drugs. And while they may be part of the distant future, can we improve clinical work in the near future?
Over the next two weeks, we will mull the future of psychiatry in terms of practice and measurement-based care. Measurement-based care has been defined simply by Scott and Lewis as “practice of basing clinical care on client data collected throughout treatment.”
This week, measurement-based care.
Next week, the end of the art of care?
While these two Readings were published in two different journals, they seek to address the future of the field, perhaps in somewhat contrasting ways.
This week, we look at a new paper published in the Psychiatric Services that offers a review of measurement-based care studies.
Measurements and Practice
“A Tipping Point for Measurement-Based Care”
John C. Fortney, Jürgen Unützer, Glenda Wrenn, Jeffrey M. Pyne, G. Richard Smith, Michael Schoenbaum, Henry T. Harbin
Psychiatric Services, 1 September 2016, In Advance
Across a wide range of treatment settings, there is a substantial gap between the outcomes achieved in randomized controlled trials and in routine mental health care. One of the main contributors to enhanced outcomes in randomized controlled trials is that treatment protocols include systematic measurement of symptom severity, followed by algorithm-based treatment adjustments when patients are not responding to care.
Although there are numerous brief, validated symptom rating scales that reliably measure change in severity of symptoms over time, only 17.9% of psychiatrists and 11.1% of psychologists in the United States routinely administer symptom rating scales to their patients. On the basis of clinical judgment alone, mental health providers detect deterioration for only 21.4% of their patients who experience increased symptom severity. Detection rates are even worse for patients whose symptoms are not deteriorating but who also are not improving as expected. The failure to detect patients who are not responding to treatment contributes to clinical inertia (defined as not changing the treatment plan despite a lack of substantial improvement in symptom severity). The use of symptom rating scales to monitor outcomes helps prompt clinicians to overcome treatment inertia and change the treatment plan when patients are not responding to treatment.
So opens a narrative review by Fortney et al., published in Psychiatric Services.
Here’s what they did:
· The authors identified articles though PubMed and Google Scholar.
· They found additional papers through citations.
· Experts were assembled in a focus group conducted by the Kennedy Forum.
Here’s what they found:
· There were 51 articles found and reviewed. The authors looked at several aspects of measurement-based care (MBC).
· Symptom rating scales. The authors noted that symptom rating scales are often used in drug trials. They suggested the possibility of doing this in regular clinical work; they noted the simplicity of the PHQ-9 (the Patient Health Questionnaire) – just nine questions on depression.
· Primary clinical benefits. They noted several advantages of MBC, including the detection of residual symptoms (which increase risk of relapse) and the enhancement of collaboration among providers.
· Empirical evidence for MBC. They noted the evidence that “frequent and timely feedback” resulted in “significantly improved outcomes.” They cited a meta-analysis involving 6 papers (6,000 patients and 300 therapists): “those randomly assigned to MBC had significantly and substantially better outcomes.” In another meta-analysis, the authors looked at 12 papers and found a “small but significant effect.” Other papers were also cited.
· Ineffective measurement approaches. The authors noted that: “not all approaches to structured assessment and feedback improve outcomes.” One-off use of scales, as an example, doesn’t seem to help. They cited a paper considering patients from an eating disorder clinic with a mid-treatment scale, finding no difference in outcomes.
· Feasibility of MBC. The authors noted large-scale use of MBC in some studies (like STAR*D) and argue that: “MBC is highly acceptable to patients.”
The authors also looked at health systems that have embraced measurement-based care. For example, at Kaiser Permanente, a stepped-care model for depression is used, where patients periodically do the PHQ-9, with the scores added to their electronic medical records, helping to judge symptom reduction and remission, and thus guiding treatment. Kaiser Permanente clinicians also use other metrics, such as distress scores, covering alcohol use and drug abuse.
The authors conclude:
For synergistic reasons, MBC may be at a tipping point in the field of mental health. There are now numerous brief structured symptom rating scales, many in the public domain, that have strong psychometric properties and that have been validated in diverse patient populations.
Technological innovations (for example, handheld devices and electronic health records) have increased the efficiency of routinely collecting symptom severity data from patients and feeding the data back to providers during the clinical encounter. There is mounting empirical evidence from trials that both pharmacotherapy and psychotherapy patients randomly assigned to MBC have better outcomes than patients randomly assigned to usual care. There is evidence from large pragmatic trials and clinical demonstration projects that MBC is acceptable to both patients and providers. There is also growing consensus from accreditation organizations, purchasers, and payers that MBC should be incorporated into performance measures and payment reforms.
A few thoughts:
1. This is a good and thoughtful review of the literature.
2. The review is a narrative review – meaning that it considers important papers under the banner of measurement-based care. The authors didn’t do a systematic review. The resulting paper provides a nice ‘waterfront’ view of the literature; it doesn’t compare effect sizes or provide a more rigorous consideration of that literature.
3. Measurement-based care is increasingly studied. In a past Reading, we looked at the Guo et al. paper from The American Journal of Psychiatry. (I consider this one of the most important papers of 2015.) Guo et al. compared the treatment of depression with two groups: those treated by psychiatrists and those treated with a depression algorithm based on scales. The latter group responded better than the psychiatrist treatment group (86.9% vs. 62.7%); they had a much higher remission rate (73.8% vs. 28.8%); their time to response to treatment was shorter (4.5 weeks vs. 8.1 weeks). See the figures below.
4. As we consider ways to improve outcomes and improve access to care, the Fortney et al. paper is worth thinking about; measurement-based care points a way forward.
5. Of course, there is a difference between the papers in the journals and the work in our clinics. Can we practically incorporate scales into practice? Tying back to the Fortney et al., Kaiser Permanente – whose enrollment is roughly equal to the population of all four of Canada’s western provinces combined – uses the PHQ-9 for patients with depression. Couldn’t this scale be use in clinics north of the 49th parallel, too?
6. But is psychiatry at risk of becoming too technical? Next week, we consider an editorial from The British Journal of Psychiatry.
Reading of the Week. Every week I pick articles and papers from the world of Psychiatry.