Despite the fact that clinicians are busy people we need to evaluate our work (writes Professor Barbara Wilson)

There is no point doing rehabilitation if it is ineffective. Research should not be regarded as something extra that is added on to our work, but should be seen an essential part of clinical practice. For every patient or client we see, we should ask ourselves “Is this patient changing and, if so, is the change due to what we are doing (or have done) or would it have happened anyway?”

All clinicians plan their treatment sessions and make notes afterwards. Based on these, we can plan and measure our interventions in such an ordered way that we are in effect conducting research. We need to begin with a question that is posed in such a way that it can be answered. Just as we do not ask general questions about medicine, surgery or pharmacology that are so broad that they cannot be answered (e.g.“Does medicine work?” or ‘Do drugs work?’), we should not pose a question such as: ‘Does rehabilitation work?’. We need to make our evaluation questions more specific, such as “Do people learn better when prevented from making mistakes during learning?”

There is no one single right way to conduct research. Methodology will depend on the questions to be answered, the facilities available, time obtainable, ethical considerations and so forth. With regard to procedures, we can use surveys, observations or experiments. A survey might be used to establish the size and extent of a question such as “what percentage of TBI patients are vegetative after a year?“ Observations might be carried out to determine behaviours, for example, “What behaviours emerge first as people recover from coma?” Surveys are more concerned with numbers while observations are more to do with what happens. The third type of research involves experiments. These are tests or trials set up to demonstrate a fact or test a hypothesis. Such experiments can be carried out with any number of groups or with individuals.

There are limitations to group studies in rehabilitation. Take a question such as “How many subjects improved?” This question applies to groups, results apply to groups and we therefore need a group design to answer the question.

On the other hand, if we ask “Is this patient improving and, if so, is the change because of our intervention or would it have happened anyway”? we are concerned with an individual and we cannot answer this with a group study. This is where single case experimental designs (SCEDs) are helpful. These are not single case reports. They are experiments which allow us to tease out the effects of treatment from the effects of spontaneous recovery and other non-specific factors.

If we wish to find out whether a particular patient is benefiting from a specific kind of procedure we need to employ a SCED. If we want to find out how many people are benefiting from this procedure we would conduct a group study. Large group studies need many people to share out individual differences while SCEDs do not have to concern themselves with this as each subject is his/her own control. Baselines are used instead of control groups.

The main types of SCEDs are reversal designs and multiple baseline designs. As a clinical psychologist I have been using SCEDS in my clinical work since the late 1970s. For further details on this important way of evaluating our clinical effectiveness there is a new book by Tate and Perdices (2018) entitled “Single Case Designs for Clinical Research and Neurorehabilitation Settings “published by Routledge.

Find out more about the work of Professor Barbara Wilson