Case misclassification in studies of spinal manipulation and arterial dissection

Area of Science:

• Neurology
• Public Health
• Epidemiology

Background:

• Spinal manipulation has been linked to cervical arterial dissection and stroke.
• Previous studies questioned a causal relationship due to potential case misclassification.
• Hypothesized underestimation of association strength and age-related differences in misclassification.

Purpose of the Study:

• To investigate the association between spinal manipulation and cervical artery dissection.
• To determine if prior studies underestimated the association due to case misclassification.
• To explore age-related differences in the association.

Main Methods:

• Identified potential cases using ICD-9 codes in the Veterans Health Administration database.
• Screened electronic medical records for "dissection" and reviewed by a neurologist.
• Determined the presence of atraumatic dissection through medical record review.

Main Results:

• Confirmed cervical artery dissection in 11.2% of cases identified by ICD-9 codes.
• Positive predictive value for dissection was higher in patients under 45 (41%) vs. 45+ (9%).
• Recalculated odds ratio for spinal manipulation and dissection was 2.15 (vs. 1.12 previously); for patients under 45, it was 6.91.

Conclusions:

• Prior studies likely misclassified cases, leading to dismissal of a causal link.
• The association between spinal manipulation and cervical artery dissection is stronger than previously reported.
• Age is a significant factor in the accuracy of case identification and the strength of the association.