Discussion
One Canadian study (Tator et al., 1987) prospectively enrolled individuals into observational, non-randomized cohorts of surgical or nonsurgical treatment for traumatic SCI. This study reported a decreased mortality in the surgically treated individuals, although neurological recovery was not different. Of note, the non-surgical group comprised significantly more cervical injuries; this is unsurprising as cervical injuries are more amenable to closed decompression and immobilization, compared to more caudal injuries. This possibly influenced the main findings including increased mortality in the nonsurgical group may be explained by the higher proportion of cervically injured individuals, many of whom died of respiratory failure, as might be anticipated. Furthermore, the surgical technology of the 1970s cannot be extrapolated to modern day medical practice, which implies that surgically treated individuals in the modern era may have better outcomes than reported in this study.
Conclusion
There is level 3 evidence (based on one cohort study; (Tator et al., 1987) that surgical treatment results in lower mortality but equivalent neurological outcome compared to nonsurgical treatment; this analysis is from an early surgical era (1970s) and should be interpreted with significant caution.
Compared to patients treated without surgery, those receiving surgery post SCI experienced lower mortality but no difference in neurological outcome; however, the techniques are from an early surgical era and should be interpreted with caution.