Many challenges persist in measuring complex domains of health status in disability. These have been acknowledged for decades and measurement tools continue to be refined. For many years, evidence-based practice in healthcare has rightly been considered best practice, but finding the research evidence often leads back to the question of how to measure outcomes and what to measure. ‘Tests or measures in clinical medicine or the social sciences can be used for three purposes: discriminating between patients, predicting either prognosis or the results of some other test, and evaluating change over time.(p. 27)
Hassani et al. are to be congratulated for choosing the third purpose and studying measures to evaluate change over time. This is important for determining if interventions ‘make a difference’ to the child or to illustrate decline in ability that supports prophylactic or preventative intervention. The study by Hassani et al. addressed questions about the importance of having sufficient data on typical performance values and has added to the knowledge by determining minimal clinically important difference (MCID) values for comparison purposes. The prospective, longitudinal design of the study is commendable. Indeed, it was surprising to find in the initial study of the Timed Up and Go test modified for children (examined in a typically developing child population), just how stable the times were compared with comparison groups of older people in previous studies.
There have been welcome changes in the knowledge and perception of disability and its classification and these have had a positive impact on intervention choices and decision-making for children with cerebral palsy (CP). These changes include the International Classification of Function, Disability and Health (ICF), improvements in diagnosis (Kirschner and Guyatt’s first purpose – discrimination between patients), and classification (the second purpose – predicting prognosis).
Choices and guidelines for decision-making remain problematic in achieving the third purpose of Kirschner and Guyatt's framework because of (1) the challenges in evaluating change over time in a complex domain for health status, and (2) the reality that clinicians work with heterogeneous population groups that are concurrently growing and developing.
A recent systematic review illustrates this point. Novak et al., described the best available intervention evidence for children with CP using a ‘traffic light’ approach. One of their conclusions was that no intervention was shown to work conclusively at more than one level of the ICF and, since there was a lack of efficacy for a large proportion of the interventions in use within standard care, more research using rigorous designs is urgently needed. Researchers must, therefore, return to the question of what should be measured in questioning the efficacy of interventions.
The two meaningful, practical, and inexpensive tests used in Hassani et al., the One-Minute Walk and modified Timed Up and Go (mTUG) tests, are not only important for being meaningful, practical and inexpensive, but they are also easily understood by the children and their carers. Performance times can become goals and therapy tasks in themselves. It is important to acknowledge the integrated skills these tests are measuring and that interventions need to be directed towards improving these skills. Also, mTUG is not necessarily a measure of disability but a measure of discrimination. An example would be in the classroom: if it takes four times longer for a child with CP to ambulate to collect a book from a shelf and return to their seat than a non-disabled child, this provides evidence that resources should be directed to providing assistance in the classroom environment to maximise education opportunities for the child.
The tools studied by Hassani et al. have an important place in being accessible to the clinician. Also significant however, is the fact that these methods of quantifying performance outcomes have the potential to reduce the gap between well-resourced tertiary health services (such as those with a gait laboratory) and settings in rural, remote or low-resourced areas. As such, they address the conundrum of choosing objective evaluative measures with good psychometric properties against the environmental or ecological approach with emphasis on process not product in evaluating the effectiveness of interventions.
Measurement of the spontaneous activity of children in their natural environment has been considered a benchmark for assessment. Prospective longitudinal studies using this approach could be a main challenge for future researchers measuring the effectiveness of interventions for children with CP.