Associations between cervical artery dissection and spinal manipulation

Cervical artery dissection (CeAD), which includes both vertebral artery dissection (VAD) and carotid artery dissection (CAD), is the most serious safety concern associated with cervical spinal manipulation (CSM). This study evaluated the association between CSM and CeAD among US adults.

Through analysis of health claims data, the researchers employed a case-control study with matched controls, a case-control design in which controls were diagnosed with ischemic stroke, and a case-crossover design in which recent exposures were compared to exposures in the same case that occurred 6-7 months earlier. The researchers evaluated the association between CeAD and the 3-level exposure, CSM versus office visit for medical evaluation and management (E&M) versus neither, with E&M set as the referent group.

2337 VAD cases and 2916 CAD cases were identified. Compared to population controls, VAD cases were 0.17 (95% CI 0.09 to 0.32) times as likely to have received CSM in the previous week as compared to E&M. In other words, E&M was about 5 times more likely than CSM in the previous week in cases, relative to controls. CSM was 2.53 (95% CI 1.71 to 3.68) times as likely as E&M in the previous week among individuals with VAD than among individuals experiencing a stroke without CeAD. In the case-crossover study, CSM was 0.38 (95% CI 0.15 to 0.91) times as likely as E&M in the week before a VAD, relative to 6 months earlier. In other words, E&M was approximately 3 times more likely than CSM in the previous week in cases, relative to controls. Results for the 14-day and 30-day timeframes were similar to those at one week.

The authors concluded that, among privately insured US adults, the overall risk of CeAD is very low. Prior receipt of CSM was more likely than E&M among VAD patients as compared to stroke patients. However, for CAD patients as compared to stroke patients, as well as for both VAD and CAD patients in comparison with population controls and in case-crossover analysis, prior receipt of E&M was more likely than CSM.

What seems fairly clear from this and a previous similar analysis by the same authors is, I think, this: retrospective studies of this type can unfortunately not provide us with much reliable information about the risks of spinal manipulation. The reasons for this are manyfold, e.g.: less than exact classifications in patients’ records, or the fact that multiple types of spinal manipulations exist of which only some might be dangerous.