Clinical interpretation and use of stroke scales

Summary

No single outcome measure can describe or predict all dimensions of recovery and disability after acute stroke. Several scales have proven reliability and validity in stroke trials, including the National Institutes of Health stroke scale (NIHSS), the modified Rankin scale (mRS), the Barthel index (BI), the Glasgow outcome scale (GOS), and the stroke impact scale (SIS). Several scales have been combined in stroke trials to derive a global statistic to better define the effect of acute interventions, although this composite statistic is not clinically tenable. In practice, the NIHSS is useful for early prognostication and serial assessment, whereas the BI is useful for planning rehabilitative strategies. The mRS and GOS provide summary measures of outcome and might be most relevant to clinicians and patients considering early intervention. The SIS was designed to measure the patient's perspective on the effect of stroke. Familiarity with these scales could improve clinicians' interpretation of stroke research and their clinical decision-making.

Introduction

The application of results from stroke trials to clinical practice needs interpretation and integration of stroke-outcome measures. The major issues that relate to stroke outcome that are amenable to measurement are neurological deficit (eg, hemiparesis or aphasia), loss of ability of perform specific tasks (eg, feeding oneself or walking), loss of ability to function in normal roles and activities (eg, employment or hobbies), and quality of life. WHO developed the International Classification of Functioning, Disability, and Health to provide a standard language for the characterisation of these domains and to broadly address the concepts of medical and social disability as a composite of body structures, functions, activity, and participation.¹ These issues might be conceptually distinct, and all can affect the perception of one's health, but in practice they overlap substantially, especially after stroke. Among patients with stroke, motor and language dysfunction substantially affect all these domains simultaneously, and therefore the validity and clinical use of these sharp distinctions must be called into question. However, all must be encompassed in the assessment of recovery after stroke.

Several tools exist to measure stroke outcomes, but they are used inconsistently among trials and their relevance may not be clear to practising physicians. Moreover, no single measure fully describes or predicts all dimensions of stroke recovery and disability. A review of outcome measures used in 51 studies of acute stroke showed that 14 impairment measures, 11 activity or participation measures, one quality-of-life measure, and eight miscellaneous other measures were used.² Some of the most widely used scales are the National Institutes of Health stroke scale (NIHSS), the modified Rankin scale (mRS), the Barthel index (BI), and the stroke impact scale (SIS).3, 4, 5 Each scale is unique, and understanding the differences is critically important for both appropriate use in clinical practice and interpretation of results reported in a clinical trial. Every scale needs to have proven reliability and defined validity, and interpretation of results requires familiarity with the characteristics of the scale.

Scales that measure neurological deficits or specific body functions can be used especially well for triage and to guide acute-treatment decisions. The NIHSS, for example, is a valuable tool for initial assessments of patients with stroke in emergency departments, hospitals, or in the prehospital setting, and is predictive of subsequent resource use and long-term outcome.6, 7, 8 The mRS and BI are commonly used to assess components of disability, such as activity and participation after stroke, and can be used to guide rehabilitation plans. The SIS was designed to gain insight into the patient's perspective on the effect of stroke.⁵ Because no individual measure fits all these roles, a composite measure, such as a global statistic derived from the scores of multiple scales, has been advocated to improve assessment of the effect of acute interventions.⁹

The purpose of this review is to educate clinicians about the use and misuse of these scales in the assessment of stroke research and in practice.