Commentary on NASCIS-2.

Abstract

Release of results of a clinical study to the public press prior to dissemination in a professional journal is a substantive issue which must undergo continued scrutiny. This report analyzes the second National Acute Spinal Cord Injury Study (NASCIS-2) and its subsequent reportage. NASCIS-2 was an elaborate protocol consisting of the randomized blinded use of methylprednisolone (MP), naloxone or placebo in 487 patients within 12 hours of spinal cord injury. Analysis of the entire population failed to show a significant clinical difference among treatments although data stratification revealed a significant change in patients with a complete loss of function who received MP within 8 hours of injury. The administration of MP was associated with a 7.1% incidence of wound infection. While these results suggested clinical benefit, incomplete data and statistical descriptions reduced assertions. Discrepancies in the report included lack of radiological data or descriptions of time-dependent surgical manipulations and extent of rehabilitative therapy. Justification for the broad 8-hour stratum and multiple hypothesis testing was unclear; incomplete design details included level of statistical significance, study power, stopping rules, trial duration and odds ratio data. The hypothesis that press release prior to publication in a professional journal would result in rapid therapy remained unproven. Reasons for pre-release, including limited peer review, physician familiarity with drug and dosage, and lack of alternative treatment or treatment complications were either unjustified or problematic. Early disclosure to the press resulted in da ta misinterpretation and oversimplification, failure to recognize exclusion or appropriate legal ramifications,and lack of communication to appropriate physicians.