Prevertebral soft tissue swelling following trauma: usefulness following tube placement

Objective. Prevertebral soft tissue swelling (PVSTS) is an important marker of spinal trauma. In patients with missed or occult cervical fractures, it may be the only indication of serious injury. It has been anecdotally suggested that PVSTS is not useful following placement of a nasogastric (NGT) or endotracheal (ETT) tube because of possible iatrogenic trauma and/or soft tissue compression. Consequently, we investigated trauma patients for evidence of PVSTS, comparing radiographs performed before and after placement of NGTs and/or ETTs. Design and patients. PVSTS at the C2 and C6 levels was measured on lateral cervical spine radiographs in 102 patients with cervical spine fractures. Measurements were obtained in 66 patients before and after placement of either an NGT (23), ETT (8), or both an NGT and ETT (35). They were also obtained in 36 control patients with fractures and neither an NGT nor ETT, at presentation and after an interval temporally matched to that in the patients with tubes in place. Results. The group with ETTs showed variable changes to PVSTS (31% increase, 63% decrease, and 6% no change). The group with NGTs also showed variable changes to PVSTS (33% increase, 53% decrease, and 13% no change). The group with ETTs and NGTs similarly showed variable changes to PVSTS (25% increase, 72% decrease, 3% no change). Surprisingly, the control group showed similar temporal changes without tube placement (49% increase, 36% decrease, 13% no change). Analyses using the one-tailed F-test of the ratio of the variance of the tube to non-tube groups and the Mann-Whitney test were performed. No significant difference was found in the prevertebral soft tissues at the C2 level between those with tubes in place and the control subjects. However, at the C6 level there was statistical significance between the groups with NGTs and both NGTs and ETTs versus the non-tube groups. Probability under the F-test for the groups with NGTs and both NGTs and ETTs was 0.001 and 0.005, respectively. Under the Mann-Whitney test, P values for groups with NGTs and both NGTs and ETTs were 0.0002 and 0.0001, respectively. Conclusion. The appearance of PVSTS showed variable and unpredictable changes following ETT, NGT, or combined ETT/NGT placement at the C2 level. This appears to be an artifact of temporal changes. However, at the C6 level, the presence of PVSTS following NGT or ETT and NGT placement (but not ETT placement alone) may be an accurate indirect sign of cervical spine injury. Received: 14 September 1999 Revision requested: 30 December 1999 Revision received: 4 February 2000 Accepted: 8 February 2000