A new algorithm based on age and sensory and motor scores in four body areas may allow clinicians to predict with great accuracy which patients will be able to walk independently 1 year after spinal cord injury.
"We have developed a simple clinical prediction rule ... that can be used by physicians to counsel patients with traumatic spinal cord injury and their families during the initial phase after injury," Dr. Joost van Middendorp and his coauthors wrote in the March 4 online issue of the Lancet.
The rule, which "can predict a patient’s long-term probability of walking independently more accurately than ... the widely used AIS [American Spinal Injury Association/International Spinal Cord Society] grading system," was developed and validated in 492 trauma patients who were treated at 19 centers across Europe, according to Dr. van Middendorp of St. Radboud University Nijmegen (the Netherlands) and his coauthors (Lancet 2011 [doi:10.1016/S0140-6736(10)62276-3]).
The authors extracted data for their prediction algorithm from the European Multicenter Study on Human Spinal Cord Injury. The study included neurologic and functional status of 1,282 patients aged 18 years or older who had sustained a traumatic spinal cord injury and were treated and assessed according to each facility’s protocol at 15 days, and 1, 3, 6, and 12 months post injury.
Of the entire cohort, 118 had 6-month outcomes and 374 had 1-year outcomes available. Most of the patients were men (77%), with an average age at injury of 44 years. The AIS grade was A in 49%; B in 13%; C in 16%; and D in 22%. Of these 492 patients, 200 (41%) were able to walk independently after the injury.
The investigators’ prediction rule assumed an age cutoff point of 65 years, because studies have shown that patients aged 65 years and older have a lower functional recovery potential. They performed motor score testing; light-touch and pinprick-sensory testing; and determination of sacral function (voluntary anal contraction and sensation). In the final analysis, the authors included only the best scores at each level of the lower extremity that was being tested, and the binary (yes/no) score for sacral function.
The patients’ ability to walk independently at 1 year was measured by the Spinal Cord Independence Measure’s indoor mobility subscale. The scale ranges from 0 (requiring total assistance) to 8 (walking without any walking aids). Scores of 0-3 signified an inability to walk or a dependency on assistance, and 4-8 signified an ability to walk independently.
The investigators’ final logistic regression model, which was deemed highly accurate, included age and four neurologic predictors at 15 days after injury:
• Age of 65 or older (score of 0-1, multiplied by –10).
• Muscle grade for the quadriceps femoral muscle (score of 0-5, multiplied by 2).
• Muscle grade for the gastrocsoleus (score of 0-5, multiplied by 2).
• Light-touch sensation at spinal level L3 (score of 0-2, multiplied by 5).
• Light-touch sensation at spinal level S1 (score of 0-2, multiplied by 5).
The rule had a minimum score of –10 and a maximum score of 40. For patients with neurologic exams that were performed within the first 15 days after injury, the probability of independent walking rose as the total score increased, such that at –10 points there was no chance of independent walking at 1 year; at 10 points, there was 35% probability; at 15 points, about 65% probability; at 20 points, almost 90% probability; and at 30 points and greater, nearly 100% probability. The area under the receiver operating characteristic (ROC) curve for the prediction rule was 0.96, indicating excellent ability to discriminate between patients who would walk at 1 year and those would not, the authors said.
Additional analyses found that adding the level of spinal cord injury, the timing of the neurologic examination, or pinprick-sensation testing did not significantly change its predictive ability.
The investigators validated the accuracy of the prediction in a separate group of 214 patients who underwent neurologic evaluation within 15 days of their traumatic spinal cord injury.
"The prediction rule had a clear additional clinical value for the prediction of an individual’s ability to walk independently in each of the AIS grades," particularly in grade B and C patients, Dr. van Middendorp said in an interview. "These are the grades of severity [for which] physicians can experience difficulties in prognosticating."
"Although our prediction rule is more accurate and less time consuming than the AIS grading system, to do accurate and reliable assessments of the four [neurologic tests], a physician must have experience in the physical examination of patients with traumatic spinal cord injury," the authors noted.