Is posttraumatic benign paroxysmal positional vertigo different from the idiopathic form?

BACKGROUND: Although head trauma is considered a common cause of benign paroxysmal positional vertigo (BPPV), clinical presentation and outcome of traumatic BPPV (t-BPPV) have not been systematically evaluated.

OBJECTIVES: To compare the clinical presentation, patient's response to physical treatment, and outcome of patients with t-BPPV with those with the idiopathic form (i-BBPV).

SETTING: Tertiary referral neuro-otology outpatient clinic.

METHODS: We reviewed the clinical records of 247 consecutive patients with posterior canal BPPV during the years 1997 to 2000. All patients were diagnosed using the Dix-Hallpike test and treated using the particle repositioning maneuver. Patients with an onset of positional vertigo within 3 days of well-documented head trauma were included in the t-BPPV group. The outcome was compared with the outcome of 42 patients with i-BPPV who were similarly treated and followed up.

RESULTS: Twenty-one (8.5%) of the 247 patients with BPPV fulfilled the diagnostic criteria for t-BPPV. The most common cause of head trauma was motor vehicle crash, documented in 57% of the cases; half of the patients additionally suffered from a whiplash injury. While the other causes were diverse, common falls were predominant. Only 2 of the patients involved in motor vehicle crashes experienced brief loss of consciousness. Sixty-seven percent of patients with t-BPPV required repeated physical treatments for complete resolution of signs and symptoms in comparison to 14% of patients with i-BPPV (P<.001). During a mean +/-SD follow-up of 21.7 +/- 9.7 months, 57% of t-BPPV patients and 19% of i-BPPV controls had recurrent attacks (P<.004).

CONCLUSIONS: The nature and severity of the traumas causing t-BPPV are diverse, ranging from minor head injuries to more severe head and neck trauma with brief loss of consciousness. It appears that t-BPPV is more difficult to treat than i-BPPV, and also has a greater tendency to recur.