Prospective cohort study on the predictors of fall risk in 119 patients with bilateral vestibulopathy

Abstract

Objectives: To identify predictive factors for falls in patients with bilateral vestibulopathy (BV). Specific variables contributing to the general work-up of a vestibular patient were compared between BV patients experiencing falls and those who did not.

Design: Prospective multi-centric cohort study.

Setting: Department of Otorhinolaryngology & Head and Neck Surgery at two tertiary referral centers: Antwerp University Hospital and Maastricht University Medical Center.

Participants: In total, 119 BV patients were included. BV diagnosis was defined in accordance with the diagnostic BV criteria, established by the Bárány Society in 2017.

Main outcome measures: Patients were divided into fallers and non-fallers, depending on the experience of one or more falls in the preceding 12 months. Residual vestibular function on caloric testing, rotatory chair testing, video head impulse test (vHIT) and cervical vestibular evoked myogenic potentials (cVEMP) was evaluated as a predictive factor for falls. Furthermore, hearing function (speech perception in noise (SPIN)), sound localization performance, etiology, disease duration, sport practice, scores on the Dizziness Handicap Inventory (DHI) and the Oscillopsia Severity Questionnaire (OSQ) were compared between fallers and non-fallers.

Results: Forty-five (39%) patients reported falls. In a sub-analysis in the patients recruited at UZA (n = 69), 20% experienced three or more falls and three patients (4%) suffered from severe fall-related injuries. The DHI score and the OSQ score were significantly higher in fallers. Residual vestibular function, SPIN, sound localization performance, etiology, disease duration, age and sport practice did not differ between fallers and non-fallers.

Conclusions: Falls and (severe) fall-related injuries are frequent among BV patients. A DHI score > 47 and an OSQ score > 27.5 might be indicative for BV patients at risk for falls, with a sensitivity of 70% and specificity of 60%. Residual vestibular function captured by single vestibular tests (vHIT, calorics, rotatory chair, cVEMP) or by overall vestibular function defined as the number of impaired vestibular sensors are not suitable to distinguish fallers and non-fallers in a BV population.

Fig 1. Sound localization test set-up.

Seven Broadband Fostex 6301 loudspeakers at intervals of 30°, located in a frontal horizontal semicircle at the subject’s head level.

Fig 2. Saccular function assessed by the cervical VEMP.

Fig 3. Overall vestibular function, defined as the number of impaired vestibular tests.

There was not a statistically significant (p = 0.054) lower vestibular function in the non-faller group. Included vestibular tests were video head impulse test of all canals (impaired if gain ≤ 0.6), rotatory chair test (impaired if gain ≤ 0.1), calorics in each ear (impaired if sum max. SPV ≤ 6°/s) and cVEMP (impaired if absent response). A number of 12 impaired vestibular tests thus represents the least residual vestibular function.

Fig 4. Sound localization test results in best-aided situation.

The root mean square localization error is a measurement of patient’s accuracy of sound localization: the higher, the less accurate sound localization. Patients who did not fall in the last year did not have significant better sound localization test results.

Fig 5. Distribution of etiologies in fallers-group and non-fallers group.

‘Other’ includes Menière’s disease, headtrauma, metabolic, auto-immune, neoplasma and other genetic disorders besides DNFA9 disease.