Key points are not available for this paper at this time.
Design
Review
Answers are grounded in this paper and cite their sources.
Urges caution interpreting nonrandomized ICD data due to blinding and mortality issues; leaves open optimal trial designs for future studies.
General modalities of analyses that have been used for ICD studies are reviewed. Published "typical" examples are briefly described. The historical cohort method is exemplified with previously unpublished data from the Seattle Cardiac Arrest Survivor database. The AVID Study database is used to compare the results obtained from nonrandomized methodologies with randomized methodologies. Particular issues related to the use of the ICD for example, mode of death, inability to blind, selection practice, and treatment decision times make this a natural pedagogic platform.
No takes yet. Share an insight, caveat, or question.
Hallstrom et al. (2000) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: