“If you can’t measure it, you can’t improve it,” said Peter Drucker, one of the greatest management consultant who ever lived.
This sentence describes the core principle around the usage of data to drive outcomes in every aspect of life, including health care (see The Bell Curve by Atul Gawande, posted at the New Yorker (credit to Prof. Jonathan Huppert for citing this in his profound talk and research done at the Hebrew University, Jerusalem).
However, is this phrase and approach apply to more holistic and complex verticals, such as art or psychotherapy? Is psychotherapy an art or science is a whole different topic I will not cover in this post, still, our belief, based on years of research and Randomized Control Trials (RCTs) conducted by some of the top researchers in the world, show that data, measurement, and feedback are essential to improve mental health treatment outcomes.
How can we even measure mental health? what does that indicate?
Healthcare treatment is heavily relied on clinician’s knowledge, experience and expertise. This is especially evident in mental health as more often than not it is not possible to see the physical symptoms of depression/anxiety (for example). Based on this assumption, measurement in mental health is not aimed to replace clinician’s evaluation but to help enrich his knowledge regarding his client state of mind at a specific point in time (usually between sessions), I often think of it as the equivalent of remotely checking the client’s blood-pressure before entering to see his primary care physician.
These measurements are also called Patient-Reported Symptom Rating Scales or Patient-Reported Outcomes Measures. They are structured instruments (in the form of questioners) that clients use to report their perceptions about the frequency and/or severity of the mental symptoms they are experiencing (Fixing Behavioral Health Care in America, The Kennedy Forum).
There are specific, clinically validated, outcome measures for various mental conditions such as for depression (PHQ-9), anxiety (GAD-7) and more (including bipolar disorders, PTSD, and schizophrenia).
Using outcome measures routinely, between every two sessions, allow clinicians to receive data that indicates the improvement/deterioration of his clients and as a result be more aligned with his client’s needs. An interesting example to the effect of such routine usage of outcome assessments can be found among clients/patients enrolled in clinical trials, these patients had much higher treatment outcomes than patients who were treated in outpatient clinics where no Patient-Reported Symptom Rating Scales were deployed (Fixing Behavioral Health Care in America, The Kennedy Forum).
The clinical evidence behind Patient-Reported Outcomes Measures
More than 15 RCTs have been conducted through the years to demonstrate the effect of outcome measures:
“Virtually all randomized controlled trials with frequent and timely feedback of patient-reported symptoms to the provider during the clinical encounter significantly improved outcomes or trended towards significance.” Fixing Behavioral Health Care in America, The Kennedy Forum
It is important to mention that these results are agnostic to the population of patients, clinical surrounding (in-patient/out-patients, etc) or the method of psychotherapy treatment (CBT/IPT/Dynamic, etc) conducted by the provider and are especially profound among clients who were not demonstrating any improvement (“not-on-track”).
As a result, both the Substance Abuse and Mental Health Services Administration, as well as the NHS, recognized these outcome measures as an evidence-based practice.
Case study: IAPT Program, NHS England
The largest and most successful implementation of outcome-based CBT is, with no doubt, the Improving Access to Psychological Therapies (IAPT) program conduct in England by the NHS and led by Prof. David M. Clark and Prof. Richard Layard.
The IAPT program began in 2008 and has transformed the treatment of adult anxiety disorders and depression in England. IAPT is widely recognized as the most ambitious program of talking therapies in the world and in the past year alone more than one million people accessed IAPT services for help to overcome their depression and anxiety, and better manage their mental health. Recently, the NHS has committed to further expand the IAPT program with an annual investment of £2.3 billion which will allow access to additional 2 million people in England (NHS Mental Health Implementation Plan 2019/20–2023/24).
All psychotherapy sessions conducted by the IAPT are measured (97% of cases have pre and post-treatment outcome-based assessments where pre-IAPT only 38% of the session had these measures). During 2015 alone, we are talking on more than 500,000 sessions with outcome-based measures.
A partial list of IAPT’s completion rates of pre and post-treatment outcome measurements Source: https://fingertips.phe.org.uk
A partial list of IAPT’s recovery rates for depression and anxiety (53.1% nationwide). Source: https://fingertips.phe.org.uk
Based on the success of the IAPT program, this model is now being adopted and/or aimed to be adopted in other countries, such as Canada, Australia, and Israel. For further information regarding the IAPT program, see this incredible talk by Prof. David M. Clark which can be found here.
The value proposition of using outcome-based treatment
There is a clear value to all stakeholders in operating under an outcome-based treatment model:
Providers — in a world shifting towards value-based care (VBC) model, many providers have clear incentives to demonstrate the quality of care they offer to their population. In addition, new CPT codes allow direct reimbursement for every outcome measures filled by clients. There is an ever-growing competition between providers (physical as well as telehealth), those who are already using outcome-based treatment have a somewhat of an unfair advantage as they can show how the care they are providing is outperforming the competition and if results are not staggering, how they intend to act in order to improve it. Needless to say, using measurement and feedback also helps on the administrator level to understand the level of improvement across his clients and his clinicians.
Payers — today mental health care is somewhat of a ‘black box’ for payers, with no quality assurance to help facilitate data-driven actions. Outcome-based measures can help payers better understand who are the top-performing providers in their network and better allocate resources to those providers. Also, the usage of routine outcome-based measures have been proven time and time again to improve treatment outcomes and potentially reduce the costs, implementing this among a costly patient population (such as chronic comorbid patients) can have an enormous effect on costs for payers.
Clients — what can be more frustrating than going to see a therapist and not receiving care that is suited for your needs? Outcome measures help therapists to be aligned with clients needs and feelings which has a direct effect as to the effectiveness of the treatment given. Outcome-based treatment empowers clients to be engaged with the course of treatment and with their feelings, it helps them to be more active and acknowledge symptoms improvement.
Hurdles for adoption and the path forward
Mostly, it’s about paperwork and integrating these inputs of data into the clinical workflow. Also, these measures only give clinicians a “sneak-peek” into their clients’ current symptoms between sessions but do not offer any explanation or feedback as to what happened in the session itself.
The usage of such measures requires high engagement rate from clients (arrive a few minutes before the session, fill out the questioners and so on), the additional workload for clinicians and do not offer any actionable insights to put to use during the upcoming session.
To create a significant change, we have to streamline the process for clients, aggregate information and drive insights which clinicians can act upon. A part of such a solution is obvious, simply digitalize part of these measures, however, this will only add an incremental benefit for the different stakeholders and will not drive significant adoption, improve outcomes or lower costs. To do so, we need to augment therapists with between-sessions as well as within-session data in an automated, frictionless manner.
To conclude, some might argue that we cannot measure what really matters and some things are just “better left unmeasured”. It might be true for other aspects of life, however, when dealing with health care and when looking at the range between zero measurement and full measurement, it is obvious we are not “full-measuring” just yet but we cannot allow ourselves to stay still and not measure at all as measurement, feedback and data is key to improve mental health outcomes.