Longitudinal changes in medical complications in adults with pediatric-onset spinal cord injury

Abstract

Objectives

To determine longitudinal changes in the occurrence of medical complications in adults with pediatric-onset spinal cord injury (SCI).

Design

Longitudinal study of long-term outcomes.

Setting

Community.

Participants

Individuals who had sustained an SCI before age 19, were 23 years of age or older at initial interview, and followed annually between 1996 and 2011. They were classified into four American Spinal Injury Association (ASIA) Impairment Scale (AIS) severity groups: C1–4 AIS ABC, C5–8 AIS ABC, T1–S5 AIS ABC, AIS D.

Outcome measures

Generalized estimating equation (GEE) models were formulated to obtain the odds ratio (OR) of having a medical complication over time.

Results

A total of 1793 interviews were conducted among 226 men and 125 women (86% Caucasian; age at baseline, 26.7 ± 3.6 years; time since injury at baseline, 12.9 ± 5.2 years). Odds of complication occurrence over time varied among severity groups, with increased ORs of severe urinary tract infection (1.05, confidence interval (CI) 1.02–1.09), autonomic dysreflexia (AD) (1.09, CI 1.05–1.14), spasticity (1.06, CI 1.01–1.11), pneumonia/respiratory failure (1.09, CI 1.03–1.16), and hypertension/cardiac disease (1.07, CI 1.01–1.15) in the C1-4 ABC group; AD (1.08, CI 1.04–1.13) and pneumonia/respiratory failure (1.09, CI 1.02–1.16) in the C5–8 ABC group; and hypertension/cardiac disease (1.08, CI 1.02–1.14) in the T1–S5 ABC group. Upper extremity joint pain had increased odds of occurrence in all injury severity groups.

Conclusion

The significantly increased odds of having medical complications over time warrants awareness of risk factors and implementation of preventive measures to avoid adverse consequences of complications and to maintain independence in individuals with pediatric-onset SCI.

Keywords:

• Longitudinal changes
• Medical complications
• Pediatric
• Spinal cord injury

Acknowledgements

The authors would like to acknowledge the support from the US Department of Education (NIDRR Advanced Rehabilitation Research Training (ARRT) Grant H133P100008) and Shriners Hospitals for Children-Chicago, Chicago, IL. The authors thank Dr Sergey Tarima, PhD, Assistant Professor, Division of Biostatistics, Medical College of Wisconsin (Milwaukee, WI) for his input and recommendations in statistical analyses.