Head shaking test and low-frequency rotation-acceleration test: comparison of sensitivity and specificity

In studying the vestibular-oculomotor reflex (VOR) in patients with equilibrium disorders, both active and passive methods are used and at different stimulation frequencies, particularly to evaluate vestibular compensation. The present study compares the sensitivity and specificity of the low-frequency pendular test and the HST in normal subjects and in subjects suffering from various vestibular pathologies but showing no signs of spontaneous nystagmus, even under optimal evaluation conditions (infrared videonystagmoscopy). These spontaneous nystagmus-free subjects underwent a routine study including: case history, infrared videonystagmoscopy (HST with 2 Hz stimulation) and ENG recording (saccadic movement, smooth pursuit, OKN, VOR, VOR-fix during pendular stimulation at 0.05 Hz and VOR after Fitzgerald-Hallpike heat stimulation). On the basis of the results the subjects were classified as follows: normal (N): case history free of any equilibrium disorders and all tests negative (122 cases; 8.1%); peripheral vestibular pathology (P): case history of equilibrium disorders and labyrinthine predominance (LP) in excess of 20% upon caloric testing (716 cases; 47.6%); central pathology (C): case history of equilibrium disorders and at least 3 pathological results from among the following tests: saccadic movement, smooth pursuit, OKN, VOR-fix (226 cases; 15.0%); mixed pathology (M): with both signs of P and C (440 cases; 29.3%). The pendular test showed signs of directional dominance (DP) higher than 10% (normal limit) in 7 cases of N (5.7%), 308 P (43.0%), 33 C (14.6%) and 162 M (36.8%). DP was higher than 10% in 55.2% of the P and M cases with onset less than 1 month before, in 42.8% of those with onset within the year and in 37.2% of those with onset more than a year before. A pathological response to the HST was observed (characterized by a series of at least 3 nystagmus shakes after a maximum latency of 15 seconds) in 0 N (0.0%), 378 P (52.8%), 4 C (1.8%) and 247 M (56.1%). The nystagmus seen was nearly always monophasic (92.5%), biphasic nystagmus was only seen in only a few cases (7.5%); moreover it was predominantly horizontal in nature (94.9%) while it was vertical in only a few of the C cases (5.1%). The HST proved pathological in 46.5% of those pathologies with onset less than 1 month before, in 55.8% of those with onset within the year and in 54.4% of those with onset more than a year before. In 213 of the cases presenting pathological response to the HST there was agreement between VOR DP at the pendular test and the direction of the nystagmus evoked by the HST: 138 P (74.6%), 0 C (0.0%), 75 M (68.2%): moreover there was no agreement in 83 cases. When the DP was lower than 10% at the pendular test, the HST proved pathological in 213 P (52.2%), 3 C (1.6%), 150 M (53.9%). Both tests gave negative results in 112 N (91.8%), 163 P (22.8%), 186 C (82.3%), 123 M (27.9%). In cases of peripheral vestibular deficit (P and M) the sensitivity of the rotoacceleration test was 40.7%, specificity 88.5%. The sensitivity of the HST was 54.1%, specificity 98.8%. The sensitivity of the association using both tests was 75.2%, specificity 85.7%.