Optimising outcome assessment to improve quality and efficiency of stroke trials

Abstract

Substantial progress has been made in treatment of stroke and much of this has been driven by large scale, multi-centre, randomised controlled trials. Although stroke is a frequent cause of mortality, stroke-related disability and functional decline is of equal or greater concern to patients and carers. Thus, to prove efficacy of an intervention for stroke, we need robust methods of describing recovery. Various functional assessment scales are available, the tool recommended as trial end point by many specialist societies and regulatory authorities is the modified Rankin Scale (mRS). We will use the mRS as exemplar to discuss contemporary research around functional assessment for stroke trials, including recent work around structured assessments, assessor training and end point adjudication panels. We will present an overview and critique of these studies and give examples where strategies to improve mRS assessment are impacting on the quality of stroke clinical trials.

Keywords::

• assessment
• clinical trial
• clinimetrics
• modified Rankin Scale
• reliability
• stroke

Financial & competing interests disclosure

K McArthur and T Quinn have participated in research in the field of post-stroke functional assessment. This work has been supported by grants from Chief Scientist Office Scotland, Chest Heart and Stroke Scotland and The Stroke Association. K McArthur and T Quinn assisted in creating training resources for mRS and BI assessment currently hosted by training campus. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Key issues

• Clinical trials in stroke require a measure of participant outcomes. As stroke is a disabling condition, measures of functional recovery are preferable to ‘hard’ clinical end points such as mortality.

• Many functional assessment scales are available, some generic and some specific to stroke. Scales are available to measure stroke recovery at levels of impairment, activity, participation and quality of life.

• An assessment scale should be assessed in terms of validity, reliability, responsiveness to change, acceptability and feasibility. There is no perfect scale and choice of outcome measure will depend on the research question, population to be studied and resource available.

• The modified Rankin Scale (mRS) is the most commonly used stroke assessment tool. It is an ordinal, hierarchical scale that grades ‘global disability’.

• A potential limitation of mRS is inter-observer variability. Poor reliability can impact on the efficiency of a trial and even modest improvements in reliability may be associated with substantial economic savings.

• Various approaches to improve reliability of mRS have been described including training and guidance in application; use of a structured assessment and use of video-recorded mRS interviews.

• Traditional dichotomous approaches to the analysis of mRS data are inefficient and techniques that use the spread of mRS are preferable.