Nature of the Non-traumatic Spinal Cord Injury Literature: A Systematic Review.

Background: Non-traumatic SCI (NTSCI) etiologies represent a markedly heterogeneous cluster of conditions defined within the consensus NTSCI taxonomy. This meta-analysis assembles evidence about the occurrence of NTSCI and its clinical outcomes with respect to 6 research domains. Purpose: To investigate the quality and quantity of clinical NTSCI evidence published in the peer reviewed literature with reference to prognosis, diagnosis, intervention, process of care, methodology, and qualitative approaches. Methods : PubMed and MEDLINE OVID MeSH heading searches were conducted for 5 common-language NTSCI descriptors. Filters were English language and Entrez date (1997-2016). Filters also controlled for case reports, editorials or errata, and invited reviews. NTSCI etiologies incorrectly classified, animal studies, and multidimensional mapping studies were excluded. Full texts were retrieved and ranked for evidence quality according to PRISMA statement guidelines, or PEDro criteria. Data were extracted and simple descriptive statistics applied. Results: The search terms non traumatic and non-traumatic SCI retrieved 282 articles, with 39 duplicates. After exclusion of 117 articles: Level 1V (60); NTSCI incorrectly evaluated (14); publication bias (2); non-English language (1); and animal experiments (1), 126 titles/abstracts were screened and ranked against criteria. Of the 8 papers allocated for full-text review, a subset of 3 articles was ranked level 1A (1) or level 11A (2); mean PEDro score 5.75±0.5. Reasons for full-text exclusions (5) were NTSCI incorrectly classified (1) and statistical limitations (4). Of the 6 domains, prognostics had adequate data yield (86) for evidence synthesis (4.8% ranked level 1A, or 11A). Notable evidence gaps were identified in qualitative (1), methodological (2), and diagnostic (8) domains. Conclusion: Therapeutic approaches require an evidence-based understanding of the distinct contexts in which NTSCI occurs, especially in less resourced settings. Our findings underscore the need for qualitative and quantitative research on the occurrence of NTSCI in all contexts.