Validity of a decision rule to reduce cervical spine radiography in elderly patients with blunt trauma

Abstract

Study objective: A decision instrument based on 5 clinical criteria has been shown to be highly sensitive in selecting patients who require cervical spine imaging after blunt trauma, while simultaneously reducing overall imaging. We examine the performance of this instrument in the elderly and explore some of the common features of geriatric cervical spine injury (CSI). Methods: The National Emergency X-radiography Utilization Study (NEXUS) was a prospective, observational, multicenter study conducted at 21 geographically diverse centers. We analyzed the performance of the NEXUS decision instrument among patients at least 65 years of age. Results: The study group consisted of 2,943 (8.6%) geriatric patients, representing 8.6% of the entire NEXUS sample. The rate of CSI was twice as great in these patients as it was in nongeriatric patients (4.59% versus 2.19%). Odontoid fractures were particularly common in geriatric patients, accounting for 20% of geriatric fractures compared with 5% of nongeriatric fractures. The frequency of patients meeting NEXUS criteria was similar in the 2 groups, with 14% of geriatric patients and 12.5% of nongeriatric patient classified as low risk. CSI occurred in only 2 low-risk geriatric patients, and these patients' injuries met our preset definition of a clinically insignificant injury. The sensitivity of the NEXUS decision instrument for clinically significant injury in the geriatric group was therefore 100% (95% confidence interval 97.1% to 100%). Conclusion: The prevalence of CSI, and especially odontoid fracture, is relatively increased among geriatric patients with blunt trauma. The NEXUS decision instrument can be applied safely to these patients, with an expected reduction in cervical imaging comparable with that achieved in nongeriatric patients. [Ann Emerg Med. 2002;40:287-293.]

Introduction

Cervical spine injury (CSI) in blunt trauma represents a classic example of the frequent use of low-yield testing to detect a high-stakes diagnosis. Physicians obtain an estimated 800,000 cervical spine radiographs for victims of blunt trauma annually, of which approximately 2.4% are positive for CSI.¹ The selective criteria proposed in the National Emergency X-radiography Utilization Study (NEXUS) were developed to safely improve the yield of cervical spine radiography. In a large, prospective, observational study involving 34,069 patients, these criteria were shown to offer a potential 12.6% reduction in cervical spine radiographs, while maintaining a sensitivity of 99.6% (95% confidence interval [CI] 98.6% to 100%) for clinically significant CSI.¹

The NEXUS decision instrument was derived and validated in a sample including patients of all ages. It is unclear whether these criteria retain their safety and efficacy in each age group. For example, geriatric patients, because of osteopenia, physical debility, and other age-related comorbid conditions, might be more vulnerable to CSI than their younger counterparts. Simultaneously, pain perception might be diminished in the elderly, which might diminish the effectiveness of individual criteria in identifying injuries. Other age-related neurologic conditions, such as dementia, could further confound interpretation of the NEXUS criteria in the elderly and significantly alter their operator characteristics and subsequent clinical utility.

This article examines the operator characteristics of the NEXUS decision instrument in the subset of geriatric patients in the NEXUS database. It also explores some of the common features of CSI found in this group.