Outcome analysis in clinical trial design for acute stroke: physicians' attitudes and choices
UMass Chan Affiliations
Department of NeurologyDocument Type
Journal ArticlePublication Date
2008-06-19Keywords
*Attitude of Health Personnel*Choice Behavior
*Clinical Trials as Topic
Comprehension
Cross-Sectional Studies
Data Interpretation, Statistical
Endpoint Determination
Female
*Health Knowledge, Attitudes, Practice
Hospitals, University
Humans
Internet
Male
Neuroprotective Agents
Patient Selection
Severity of Illness Index
Stroke
Treatment Outcome
United States
Neurology
Neuroscience and Neurobiology
Metadata
Show full item recordAbstract
BACKGROUND: Thrombolysis remains the only proven therapy to benefit acute ischemic stroke (AIS) patients. Recent studies have introduced more sensitive outcome measures such as the shift analysis to detect a treatment effect in AIS trials and are also including imaging as a surrogate of injury. METHODS: We conducted a cross-sectional, internet-based survey of academic neurologists regarding their attitudes, choices and understanding of various outcome measures in clinical trial design for AIS. The survey population consisted of neurologists who specialize in the care of stroke patients and are on faculty at university-affiliated hospitals in the USA. RESULTS: 152 of 300 neurologists completed the survey. There were 79% men and 21% women. Among commonly used outcome scales in acute stroke, the most frequent ones selected for use as trial primary endpoints were the global statistic (59%), modified Rankin scale (mRS) (52%), and NIHSS (30%). When given choices about which outcome on the mRS would justify a therapeutic intervention, 54% chose a shift analysis of change in the distribution of outcomes and 39% chose a dichotomized outcome (mRS CONCLUSION: The majority of respondents accepted an analysis of the entire distribution of the mRS scores as an appropriate endpoint analytic technique in AIS trials and did not require the traditional dichotomized outcome to demonstrate a treatment effect; however, a better understanding of the shift strategy is needed. Our data also support the importance of incorporating mismatch imaging into future neuroprotection trials.Source
Cerebrovasc Dis. 2008;26(2):156-62. Epub 2008 Jun 17. Link to article on publisher's siteDOI
10.1159/000139663Permanent Link to this Item
http://hdl.handle.net/20.500.14038/37691PubMed ID
18560219Related Resources
Link to Article in PubMedae974a485f413a2113503eed53cd6c53
10.1159/000139663