The phrase evidence-based practice is now common parlance in mental health care. The call for using evidence-based practice can be heard across psychology, social work, psychiatry, occupational health, and a range of other professions.
Often, such “evidence” consists of data from randomised controlled trials (RCTs), non-randomised trials, case studies, qualitative focus groups and interviews, and a range of other sources.
Nevertheless, many decisions in mental health care, whilst being informed by such an evidence base, rely heavily on clinical judgement. As noted in Daniel Kahneman’s seminal Thinking, Fast and Slow, clinical judgement is often not as reliable as it seems, especially for longer term outcomes.
Then, there is also a need for practice-based evidence or the collection of evidence during on-the-ground work to measure outcomes for the individual client.
Unfortunately, the simple removal of psychiatric diagnosis (e.g. “You are no longer depressed”) cannot be considered an accurate measure of change for an person. For decades the unreliability of psychiatric diagnosis has been flagged as an issue and unfortunately not much has changed. Indeed, such diagnostic categories are not even considered scientifically valid.
However, range of scales and measures are available to monitor change for various problems. One example might be the Internal States Scale for someone having problems with fluctuating mood, the Hospital Anxiety and Depression Scale for someone who is feeling low, the Self-Compassion Scale for someone who is trying to develop compassion, or power mapping to measure someone’s level of personal control within their lives. Positive aspects of a person’s life can also be measured, such as user-defined recovery, quality of life, empowerment, and subjective wellbeing to name but a few.
Outcome measures can also be more tailored to the individual. Subjective units of distress, personal goals (with a clear 1-5 scale of how close one is to achieving the goals), subjective blob trees, and six part stories are other ways to measure the impact of interventions. In fact, such personalised measures may be more meaningful than a set of standardised questionnaires, which can often lack context.
Measures can be helpful to see whether an intervention is staying on track, and may give permission to discuss topics that are difficult to bring up (such as sex drive or self-harm). It may be that the professional has missed something in their questioning, which is revealed through outcome measures to be important to the client.
Clients may also value being able to review their progress, and there are some indications that the use of outcome measures can improve therapeutic outcomes. Outcome measures may additionally support the clinician to reflect upon their work, and ensure that the clinician’s views on progress match the client’s reality.
There are, however, some concerns about using outcome measures in clinical practice. This might include burden, time and paperwork for both client and professional, concerns about relevance and helpfulness of scales, and also concerns that any outcome scores may be misused by others. An over-reliance on scales to measure outcome may not be helpful or meaningful. Additionally, scales can only see a change in the concept that they are measuring – for example, a mood scale may miss out key components in somebody’s social life.
As such, like any clinical tool, outcome measures should be used in a person-centred fashion. Good clinical practice involves firstly using an outcome measure that is relevant to the individual. If a person who feels tired and low wants to get out more and make new friends, it may be helpful to set goals related to increased activity (rather than, say, simply using a scale of how low the person feels).
Secondly, an outcome measure should be feasible for the person. For some people, an hourly record of their day might be too much to ask, whereas for someone else it may be helpful and motivating. This can be subject to some experimentation and playfulness – the measure does not have to be “right” the first time but can be considered a work-in-progress.
Thirdly, an outcome measure should gather enough information to be useful, and no more. After all, measures take time and effort to fill in, and it is not ethical to ask someone to collect information that will not be used.
Finally, the progress of outcome measures should be shared and discussed with the client. They should not be collected simply for the sake of the professional, or the service, but should be a meaningful addition to any therapeutic intervention.
Therefore, whilst the concept of evidence-based practice is well known in mental health care, there is an important role for collecting outcome measures as part of everyday practice (practice-based-evidence). As long as such measures are focussed on outcomes which are meaningful to the client, and are used in an ethical and person-centred fashion, they can prove a valuable – if not integral – part of clinical work.
