Sacral sparing in SCI: beyond the S4-S5 and anorectal examination.

Background Context: Sensory and/or motor function sparing, including the S4-S5 spinal cord segment, is central to classifying neurologic impairment after spinal cord injury (SCI) using the American Spinal Injury Association Impairment Scale (AIS) grades within the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI). Within the ISNCSCI protocol, which is essential for both clinical and research purposes, assessing sacral sparing requires an anorectal and S4-S5 examination. However, in situations where these data are incomplete, the relationships between anorectal/S4-S5 examinations and functional preservation at more rostral sacral segments may be useful.

Purpose: To evaluate whether slightly more rostral sensory and motor outcomes of the ISNCSCI can accurately predict caudal sacral sparing (S4-S5 dermatome sensation, "deep pressure" anal sensation [AS], and voluntary anal contraction [AC]).

Study Design: Retrospective analysis of the European Multicenter Study about Spinal Cord Injury database.

Patient Sample: One thousand four hundred sixty-seven AIS-A, AIS-B, and AIS-C subjects.

Outcome Measures: International Standards for Neurological Classification of Spinal Cord Injury examinations.

Methods: The value of six factors (sensory preservation at S1, S2, and S3; motor preservation at S1; motor function at more than three segments below the motor level; and sensory function at more than three segments below the neurologic level) for predicting ISNCSCI sacral sparing measures (AS, S4-S5 dermatome sensation, AC) was evaluated. Combinations of the most promising factors were then evaluated for their ability to accurately predict the AIS grade.

Results: Preserved sensation at the first sacral segment (S1S) provided good prediction (90.5%) of caudal sacral sensory sparing (ie, AS or S4-S5 sensation). Voluntary anal contraction was accurately predicted by preserved motor function within the first sacral segment (S1M) in 85.4% of cases. The alternate classification schemes evaluated for accurately predicting the AIS classification grade were S1S+S1M and S1S+motor preservation more than three segments below the motor level. The ability of these schemes to accurately predict AIS grades was stable over time but varied with the rostrocaudal level of spinal injury. For the initial baseline examination, the alternate classification schemes were accurate in ~95% of cases for T2-T9 SCI, with slightly lower accuracy for cervical SCI (~80%).

Conclusions: There are close relationships between functional sparing at different sacral segments. These relationships can be used to estimate AIS grades when complete information about the anorectal and S4-S5 examination is not available. The accuracy of the classification remains stable over time, while the increased variability in lower levels of SCI, that is, lumbar injuries, emphasizes the importance of careful sacral examinations. The highly reliable predictive values of S1-S3 segments can complement conclusions from anorectal examinations if the latter are considered to be confounded or incomplete.