Subgroup Analysis

  • NASCIS 2 only showed benefit in the subgroup initiating treatment within 8 hours, raising concerns about subgroup analyses without a benefit in the larger study group.
  • The 8-hour cutoff was chosen post-data collection and lacks physiologic support for its use.
  • Further analysis of patients in the early treatment group by severity subgroup results in small sample sizes and multiple comparisons, which lacks consistent benefit.

Unreported Data

  • Demographic and severity comparisons between treatment groups in the early treatment subgroup are unavailable, making it impossible to assess the success of randomization and potential confounding.
  • It is unclear if the treatment groups received the same medical treatment and surgical options, potentially confounding the results.
  • The decision to report only right-sided motor data while reporting bilateral sensory scores has been questioned, with the NASCIS group maintaining that the right- and left-sided scores are similar.
  • The functional significance of a small improvement on a 70-point motor scale score is unclear, leading to debate about the clinical significance of the results.

Risk of Complications

  • In NASCIS 1, there was a statistically significant increase in wound infections in the 1000-mg-daily group.
  • Two-week mortality was 3.1 times higher in the 1000-mg-daily group, and 28-day mortality was 1.92 times higher in the 1000-mg-daily group.
  • In NASCIS 2, rates of wound infections and gastrointestinal bleeding were higher in the MP group, but the differences were not statistically significant.
  • In NASCIS 3, at 6 weeks, there was a statistically significant increase in severe pneumonia in the 48-hour group compared with the 24-hour group, and there was also a nonsignificant trend to increased incidence of severe sepsis in the 48-hour group.
  • Like any randomized controlled study, the NASCIS trials were powered to demonstrate the primary outcome, not to establish safety.


  • The NASCIS trials have faced criticisms about the use of subgroup analysis, potential confounders, clinical significance of results, reproducibility, and risk of complications.
  • The Cochrane meta-analysis recommends the use of MP for acute SCI, but critics argue that more research is needed to fully understand the risks and benefits.

Practice Patterns for Steroid Use in Acute SCI

  • Steroids have been used in acute SCI treatment since the 1960s, but negative results from NASCIS 1 caused some providers to discontinue the practice
  • Studies of practice patterns in North America and the UK suggest uneven adoption of the NASCIS recommendations
  • Evidence suggests the use of MP for acute SCI may be decreasing, but controversy continues over clinical benefit
  • Medicolegal concerns are often cited as the justification for use

Steroid Use Among North American Surgeons Treating Patients with Acute SCI

  • In 1999, a survey of trauma facility medical directors in Colorado found that 98% used steroids for acute SCI, but less than half (48.7%) indicated that the evidence supported the use of steroids
  • In Canada, a 2001 survey found that 77% of respondents used steroids in acute SCI, but only 30% believed there was a benefit. In a 2006 repeat survey, only 24% reported using steroids for acute SCI
  • A 2006 survey of the North American Spine Society found that 85% of respondents used steroids in acute SCI if they could be initiated within 8 hours of the injury. Of those, 110 did not believe there was a clinical benefit but cited medicolegal concerns as a reason for use, while 65 thought steroids provided clinical benefit
YearCountryRespondents% Using Steroids% Believing in Benefit