Defining Success: Basics of Measurement Scales and Clinical Research in Neurovascular Disease
Pearls
A “clinically successful outcome” should be measured with a validated scale.
Neurologic outcome scales are typically related to function or examination.
Bias must be carefully eliminated or minimized, or it will cloud and falsify results.
The appropriate statistical test will highlight the hypothesized result. The wrong statistical test is more likely to be falsely negative (fails to detect a difference) or falsely positive (detects a difference where none exists).
Running numerous statistical tests on a large data set (“datadredging”) will be apparent to sophisticated readers and reviewers.
♦ Measurement in Neurovascular Disease
Minimizing the occurrence, impact, and long-term consequences of neurovascular disease requires an ability to measure neurologic function. Not all outcomes can, or should, be measured in all patients. In general, the more accurate a scale, the more expensive (in time, equipment, and sophistication of the examiner) it is to obtain a result.
Mortality
When patients are lost to follow-up, mortality can usually be ascertained by searching publicly available vital records. Mortality is an insufficient measure in neuroscience, however, because it does not measure function or quality of life. For many patients (and potential patients) avoiding severe disability or a vegetative state is more important than staying alive in the physiologic sense.1
Functional Scales
Functional scales attempt to measure what the patient can do. There are a variety of available metrics. Some are self-explanatory, have available on-line training, or can be administered with a standardized questionnaire.
Glasgow Coma Scale 2
The Glasgow Coma Scale (GCS, Table 5.1 ) is the most widely used neurologic examination for life-threatening neurologic disease because of its simplicity and prognostic validity. There are ceiling effects (the best possible score, 15, can be associated with disability) and floor effects (the worst possible score, 3, does not mean death).
FOUR Score 3
The Full Outline of UnResponsiveness (FOUR) score is also designed to assess comatose patients, but includes assessments of respiratory response on the ventilator and the response of pupils to light. It is more helpful for determining minimal brainstem function than the GCS.
National Institutes of Health Stroke Scale (NIHSS) 4
The possible scores on this validated neurologic exam range from 0 (no abnormality) to 42 (the worst possible score). The disadvantages of the NIHSS include its bias toward language (as opposed to visual-spatial) function,5 and its relatively crude assessment of consciousness. Comatose patients are unable to do much of the exam.
Eyes | Motor | Verbal |
6 Follows | ||
5 Localizes | 5 Oriented | |
4 Spontaneously open | 4 Withdraws | 4 Disoriented but in context |
3 Open to voice | 3 Flexor posturing | 3 Inappropriate words |
2 Open to pain | 2 Extensor posturing | 2 Unintelligible |
1 No response | 1 No response | 1 No response |
Modified Rankin Scale
The modified Rankin Scale (mRS, Table 5.2 )6 is graded from 0 (no symptoms) to 6 (death). It is heavily geared toward functional independence and the ability to walk. There is a validated questionnaire that improves inter-rater reliability.7 The score can be obtained over the phone. Serial scores can be followed over time. The mRS is not sensitive to cognitive dysfunction. An awake, interactive, but bed-bound patient scores the same as a vegetative patient.
Glasgow Outcome Scale
The Glasgow Outcome Scale (GOS) is similar to the mRS, and is graded from 1 (dead) to 5 (able to return to work or school). The major distinctions in this scale are a vegetative state, the ability to live independently, and the return to useful work. A more detailed eight-step version is the GOS-Extended (GOS-E) scale.8
Barthel Index 9
The Barthel Index focuses on the functions of daily living, including dressing, eating, and continence. Scores are from 0 (not independent in any item) to 100. It may be particularly helpful for planning rehabilitation.10
Sickness Impact Profile 11
The Sickness Impact Profile (SIP) is a 136-item questionnaire that measures functioning in a variety of domains, including function, daily living, and psychosocial health. The SIP has been validated in subarachnoid hemorrhage.
Cognitive Scales
Cognitive scales are important for determining if a conscious patient can live independently or return to productive work. Unfortunately, a substantial number of patients who survive an intensive care unit (ICU) stay will not be able to complete cognitive testing. Potential barriers include the inability to present for the test, the cost of trained staff, the unavailability of clinic and testing space, and the cognitive disability itself from neurologic disease. The U.S. National Institute of Neurological Disorders and Stroke (NINDS) and the Canadian Stroke Network have proposed standards for 5-minute, 30-minute, and 60-minute batteries of complementary tests for patients with neurovascular disease.12
Telephone Interview for Cognitive Status (TICS) 13
The TICS is a multistep cognitive exam for a variety of tasks including memory, attention, declarative knowledge, and orientation. It performs well compared with comprehensive batteries, and highlights patients who are likely to have abnormalities on more detailed testing.14
Grooved Pegboard Test
This test measures the ability of a subject to place grooved pegs in a standard board. Different results may be obtained for the dominant and nondominant hand. This test requires a cooperative patient with at least some motor skill.
Trail-Making
The primary outcome of this test is the time to connect the dots on a testing surface. Online tests are also available for remote measurement. Standard sheets are commercially available.
Mini–Mental Status Exam 15
This popular exam measures cognitive performance in a variety of domains. It is brief and meant to be done at the patient’s bedside. Brief tests of memory, cognition, visualspatial function, and language are included.
Imaging End Points
Volume
In ischemic stroke, ischemia to large portions of the middle cerebral artery territory is associated with a higher risk for hemorrhage. The Alberta Stroke Program Early CT Score (ASPECTS) is graded from the CT scan16 and correlates well with infarct volume by measuring whether discrete regions of brain tissue are hypodense. The intracerebral hemorrhage (ICH) volume and subsequent clot growth17 are important predictors of outcome. The ICH volume can be calculated with dedicated software, or roughly estimated using the abc/2 method (height times width times diameter divided by 2).18 Larger intracranial aneurysms are more difficult to obliterate and have a worse prognosis.
Cerebral Infarction
Cerebral infarction is common after a subarachnoid hemorrhage (SAH).19 Most, but not all, infarctions can be linked to a vasospasm. Cerebral infarction should at least be recorded as present or absent. Further quantifying cerebral infarction by location (cortical, subcortical, or both) and number (single or multiple)20 may be more accurate.
Atrophy
A variety of neural insults and cerebral infarction may lead to brain atrophy. Decreased brain volume correlates with progression of dementia21 and a worse performance on cognitive testing.
Angiography
Angiography is a common examination for SAH and proximal ischemic stroke. Angiographic vasospasm is often scored dichotomously (present or absent), or qualitatively (none, mild, moderate, or severe). When reporting such data, the criteria should be clearly stated (“flow-limiting,” based on a percentage change from baseline, diameter, etc.).
In acute ischemic stroke, important variables include the degree of stenosis, arterial occlusion, and the interventions that are performed. The faster and more completely that flow is restored, the better the radiographic and clinical outcome.
Combining Complementary End Points
Sometimes combining two scales is more helpful that using just one. For example, the NIHSS may clarify whether or not a normal GCS score is associated with a hemiparesis. A patient who is unable to live independently (mRS 4) may have a nearly normal neurologic exam but be impulsive or cognitively impaired.
♦ Key Web Resources for Neurologic Scales
Center for Outcome Measurement in Brain Injury (www.tbims.org): provides an introduction, online training, and many commonly used outcomes scales
Stroke Center at the University of Washington (www.strokecenter.org/trials/scales/scales-overview.htm): provides an overview of commonly used outcome scales and a registry of clinical trials in stroke
www.NIHstrokescale.org: provides online training in the NIHSS in a variety of languages
♦ Clinical Research in Neurovascular Disease
Selection of a Condition to Study
Characteristics of potentially successful clinical projects include enrolling a sufficient number of patients (a minimum of a few per month, but they need not be enrolled at all hours of the day and night), reliable ascertainment of the patient’s suitability for the study (by admission to a defined location, or enrolled by a reliable screener or by referral), clear measurement techniques, and well-trained staff working under the direction of a project leader. Examples of project topics include clinical vasospasm, severe neurotrauma, and monitoring in coma, all of which can be reasonably studied in larger medical centers. Without observational pilot data, it is difficult to judge precisely how many patients will be available. For example, ischemic stroke is common, but presentation within 3 hours is not. It is insufficient to cite an index case (“We see it”) or two (“We see it over and over”) as a foundation for research.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
