The valid confirmation of a positive change (improvement) in a patient's health status due to intervention has been at the core of medicine and rehabilitation since their very inception as clinicians always aspired to ensure that treating their patients had led to successful outcomes both in acute and chronic conditions. However what is change: either improvement or worsening (aggravation), is a complicated issue which involves clinical as well as statistical considerations. Change invariably relates to a difference in some measurable entity and almost always it relates to a time span. The confirmation of clinical change is important both for varying the treatment course (if necessary) and for the termination of treatment when the latter has reached wither its prescribed objective or a plateau. Since in the context of rehabilitation, the outcome measures (OM) are strongly linked to performance, determination of change in the latter is confounded by many factors, collectively known as the error of measurement, which render a decision regarding clinically meaningful change, highly involved. This is further complicated by the stability of the observed OM, the so-called reproducibility of the OM, and the accuracy of the measurement instrument. The higher the reproducibility the lower is the error. Moreover, in order to proclaim change, in most cases a positive one, it is necessary for the difference in outcome scores (i.e. the change) to surpass the error of measurement, in varying degree of rigor. This paper describes selected methods associated with determination of change and focuses predominantly on the difference between a simple difference in scores ('simple change'), a significant difference in scores and the so-called clinically meaningful change in scores which is considered today as the benchmark for confirmation of a real change.
Keywords: Change; Clinically meaningful; Error; Reproducibility.