Challenges in Stroke Clinical Trials: Improving Outcome Measurements

Stroke remains one of the leading causes of disability and mortality worldwide. Despite advances in treatment options, developing new therapies faces substantial hurdles.  

Clinical trials in stroke present unique challenges ranging from patient heterogeneity to the complex nature of measuring meaningful outcomes. This blog explores these challenges, focusing on outcome measures such as the Barthel Index, the National Institutes of Health Stroke Scale (NIHSS), and the Modified Rankin Scale (mRS), highlighting innovations like the structured interview for the mRS and Signant Health's PureSignal Analytics that could improve stroke trials. 

The Challenge of Measuring Recovery 

Stroke affects patients in multidimensional ways—physically, cognitively, and emotionally—necessitating multifaceted outcome measures. The ideal measure should be valid, reliable, sensitive to meaningful changes, easy to administer, applicable across different stroke types, and reflective of functional independence.  

No single measure perfectly addresses all these requirements, which is why trials typically employ multiple complementary measures. 

Traditional Outcome Measures and Their Limitations 

The Modified Rankin Scale: Current Standard with Persistent Problems 

The mRS provides a global assessment of disability on a scale from 0 (no symptoms) to 6 (death). Its popularity stems from its simplicity, focus on overall disability, and correlation with other measures. However, traditional assessment suffers from: 

• Subjectivity: Without structured guidance, assessors interpret categories differently 
• Cultural and Contextual Insensitivity: Doesn't account for pre-stroke lifestyle or goals 
• Limited Granularity: May miss improvements that don't cross category thresholds 
• Recency Bias: Undue influence from the patient's condition on assessment day 

The Structured Interview for the Modified Rankin Scale 

The structured interview for the mRS (SI-mRS) uses a standardized questionnaire to guide assessors through the rating process. Benefits include: 

• Improved Reliability: Studies show approximately 25% better inter-rater reliability 
• Reduced Assessment Bias: Ensures all patients are evaluated on the same criteria 
• Increased Precision: More consistent categorization of patients 
• Enhanced Training Efficiency: More straightforward training across trial sites 
• Better Detection of Treatment Effects: Reduced noise allows smaller sample sizes 

This structured interview has been shown to enhance interobserver reliability¹.

NIHSS: Quantifying Neurological Deficits 

The NIHSS has become the gold standard for assessing neurological impairment, offering standardized administration, excellent inter-rater reliability, and strong predictive value. However, it also has limitations: 

• Bias Toward Anterior Circulation Strokes: Potentially underestimates posterior circulation events 
• Limited Sensitivity to Recovery: Less sensitive to functional improvements during recovery 
• Training Requirements: Resource-intensive in large multicenter trials 
• Disconnect from Functional Impact: Scores may not correlate with real-world functioning 

The Barthel Index: Measuring Activities of Daily Living (ADLs) 

The Barthel Index evaluates a patient's independence in performing activities of daily living across ten domains. Despite its widespread use, it presents several challenges: 

• Ceiling Effect: Many patients with significant disabilities may achieve maximum scores 
• Insensitivity to Cognitive Impairment: Focus on physical abilities misses cognitive deficits 
• Nonlinear Recovery Patterns: Assumes all ADLs have equal weight and difficulty 
• Administrative Variability: Different administration methods yield different results

Leveraging Signant Health's Expertise 

With decades of experience in clinical trial optimization, Signant Health brings specialized knowledge to address the unique challenges facing stroke trials. Their comprehensive approach transforms traditional assessment methods through:

Enhanced Training and Certification 

• Implements standardized administration of scales 
• Provides training modules focusing on common challenges 
• Provides periodic recalibration activities

PureSignal Analytics 

Technology solutions are increasingly crucial in addressing persistent challenges in stroke trials. Signant Health's PureSignal Analytics enhances data quality throughout the trial process through: 

Real-time Quality Monitoring 

• Can begin as soon as data starts to be acquired from the study
• Identifies patterns of inconsistent scoring between raters and sites 
• Detects rater drift over time 
• Flags in-between scales. For example, identifying when different scales show conflicting results (e.g., mRS and Barthel Index)
• Spots unusual scoring patterns: Detecting statistical outliers compared to other participants/sites 
• Detects participant non-compliance with eDiaries

Data-Driven Decision Making 

• Enables early identification of sites with potential data quality issues 
• Quantifies different sources of variability in assessments 

Integration of Multiple Outcome Measures 

• Identifies inconsistencies between related measures 
• Incorporates patient-reported data alongside clinician assessments 
  

Future Directions 

Researchers continue exploring additional enhancements to outcome assessment, including: 

• Digital tracking through wearable devices and external sensors 
• Domain-specific assessment batteries 
• Real-world functional performance observation 
• Machine learning approaches for composite endpoints

Conclusion 

The challenges in stroke clinical trials extend beyond outcome measurement to patient heterogeneity and stroke pathophysiology complexity. However, improvements in outcome measurement represent a critical step toward more efficient, reliable trials. 

The structured interview approach to the mRS, combined with advanced data analytics platforms like Signant Health's, can reduce variability and increase precision in primary outcome measures. This allows for trials requiring fewer patients while reliably identifying beneficial interventions. 

As we refine our approaches to stroke outcome assessment, patients will gain access to effective therapies more quickly. The evolution from basic scales through structured approaches enhanced by sophisticated analytics represents our growing sophistication in capturing the complex reality of stroke recovery. 

Future trials should incorporate both structured assessment approaches and technology solutions for quality control. Through these combined methodological and technological improvements, we may finally overcome persistent challenges in stroke treatment development. 

About the Authors

Pascal Goetghebeur is an experienced CNS pharmacologist and behavioral scientist specializing in psychiatry and neurodegeneration. At Signant Health, he serves as Clinical Principal for Cognition, Science & Medicine in the Digital Health Sciences division, focusing on the role of cognition in drug development and managing the clinical aspect of Signant’s proprietary cognitive assessment battery.

Lewis M. Fredane, MD, is the Clinical Vice President at Signant Health, serving as the Therapeutic Area Leader for neurology. He oversees Signant’s eCOA, Rater Training, and Quality Assurance in this field. Previously, Dr. Fredane was a practicing neurologist and clinical assistant professor at Drexel University College of Medicine. He has experience in pharmaceutical drug development with companies like Sanofi/Genzyme and Meda Pharmaceuticals and has worked with CRO Omnicare Clinical Research.