A Set of Eight Key Questions Helps to Classify Common Vestibular Disorders-Results From the DizzyReg Patient Registry

Patients’ Experiences with Management of Benign Paroxysmal Positional Vertigo: Insights from the Vestibular Disorders Association Registry

Introduction

Benign paroxysmal positional vertigo (BPPV) is the most frequently occurring peripheral vestibular disorder. Clinical practice guidelines (CPG) for BPPV exist; however, little is known about how affected patients perceive their condition is being managed. We aimed to leverage registry data to evaluate how adults who report BPPV are managed.

Material and Methods

We retrospectively analyzed of data from 1,262 adults (58.4 ± 12.6 years old, 81.1% female, 91.1% White) who were enrolled in the Vestibular Disorders Association Registry from 2014 to 2020. The following patient-reported outcomes were analyzed by proportions for those who did and did not report BPPV: symptoms experienced, falls reported, diagnostics undertaken, interventions received (eg, canalith repositioning maneuvers [CRMs], medications), and responses to interventions.

Results

Of the 1,262 adults included, 26% reported being diagnosed with BPPV. Many adults who reported BPPV (83%) also endorsed receiving additional vestibular diagnoses or may have had atypical BPPV. Those with BPPV underwent magnetic resonance imaging and were prescribed medications more frequently than those without BPPV (76% vs 57% [χ²=36.51, p<0.001] and 85% vs 78% [χ²=5.60, p=0.018], respectively). Falls were experienced by similar proportions of adults with and without BPPV (55% vs 56% [χ²⁼=11.26, p=0.59]). Adults with BPPV received CRMs more often than those without BPPV (86% vs 48%, χ²=127.23, p<0.001). More registrants with BPPV also endorsed benefit from CRMs compared to those without BPPV (51% vs 12% [χ²=105.30, p<0.001]).

Discussion

In this registry, BPPV was often reported with other vestibular disorders. Healthcare utilization was higher than would be expected with care based on the CPG. The rates of falls in those with and without BPPV are higher than previously reported. Adults with BPPV reported significant differences in how their care is managed and their overall outcomes compared to those without BPPV.

Conclusion

Patient-reported outcomes provide useful information regarding the lived experience of adults with BPPV.

Validation of a comprehensive diagnostic algorithm for patients with acute vertigo and dizziness

BACKGROUND

Vertigo and dizziness are common complaints in emergency departments and primary care, which pose major diagnostic challenges due to various underlying etiologies. Most supportive diagnostic algorithms concentrate on either identifying cerebrovascular events or diagnosing specific vestibular disorders or are restricted to specific patient subgroups.

METHODS

The study was conducted in the scope of the 'PoiSe' project (prevention, online feedback, and interdisciplinary therapy of acute vestibular syndromes by e-health). A three-level algorithm was developed according to international guidelines and scientific evidence addressing both, the detection of cerebrovascular events and the classification to non-vascular vestibular disorders (unilateral vestibulopathy, benign paroxysmal positional vertigo, vestibular paroxysmia, Menière's disease, vestibular migraine, functional dizziness). The algorithm was validated on a prospectively collected dataset of 407 patients with acute vertigo and dizziness presenting to the emergency department at LMU Munich.

RESULTS

The algorithm assigned 287 of 407 patients to the correct diagnosis, corresponding to an overall accuracy of 71%. Cerebrovascular events were identified with high sensitivity of 94%. The six most common vestibular disorders were classified with high specificity above 95%. Random forest identified the presence of a paresis, sensory loss, central ocular motor and vestibular signs (HINTS), and older age as the most important variables indicating a cerebrovascular event.

CONCLUSIONS

The proposed diagnostic algorithm can correctly classify the most common vestibular disorders based on a comprehensive set of key questions and clinical examinations. It is easily applied, not limited to subgroups, and might therefore be transferred to broad clinical settings such as primary health care.

Use of Neuroimaging for Patients With Dizziness Who Present to Outpatient Clinics vs Emergency Departments in the US

Importance

Overuse of costly neuroimaging technology is associated with low-value care for the prevalent symptom of dizziness. Although quality improvement initiatives have focused on the overuse of computed tomography (CT) scans in emergency departments (EDs), most patients with dizziness present to outpatient clinics. To inform practice and policy, a comprehensive understanding of the uses and costs of neuroimaging across settings and episodes of care is needed.

Objective

To characterize neuroimaging use, timing, and spending as well as factors associated with imaging acquisition within 6 months of presentation for dizziness in outpatient vs ED settings.

Design, Setting, and Participants

This cross-sectional study of commercial and Medicare Advantage claims for 805 454 adults (≥18 years of age) with new diagnoses of dizziness was conducted from January 1, 2006, through December 31, 2015. Data were analyzed from October 1, 2020, to September 30, 2021.

Main Outcomes and Measures

Use of neuroimaging (CT scan, magnetic resonance imaging [MRI], angiography, and ultrasonography) and total spending on neuroimaging were measured. Kaplan-Meier analysis was performed. The associations of neuroimaging with setting, sociodemographic characteristics, and clinicians were estimated with multivariable analyses.

Results

A total of 805 454 individuals with dizziness (502 055 women [62%]; median age, 52 years [range, 18-87 years]) were included in this study; 156 969 (20%) underwent neuroimaging within 6 months of presentation (65 738 of 185 338 [36%] presented to EDs and 91 231 of 620 116 [15%] presented to outpatient clinics). The median time to neuroimaging was 0 days (95% CI, 0-2 days) after ED presentation and 10 days (95% CI, 9-10 days) after outpatient presentation. Neuroimaging was independently associated with advanced age, comorbidity, race and ethnicity, ED presentation, and outpatient clinician specialty. Across sites, a head CT scan was the most used test on presentation date (92% of tests [46 852 of 51 022]). Within 6 months of presentation, a head CT scan was the most used test (47% of all tests [177 949 of 376 149]), followed by brain MRI (25% [93 130 of 376 149]), cerebrovascular ultrasonography (15% [56 175 of 376 149]), and magnetic resonance angiography (9% [34 026 of 376 149]). Of $88 646 047.03 in total neuroimaging spending, MRI accounted for 70% ($61 730 251.95), CT scans for 19% ($16 910 506.24), and ultrasonography for 11% ($10 005 288.84). Per-test median spending ranged from $68.97 (CT scan of the head) to $319.63 (MRI of the brain) among commercially insured individuals and $43.21 (CT scan of the head) to $362.02 (MRI of the orbit, face, and neck) among Medicare Advantage beneficiaries.

Conclusions and Relevance

The findings of this cross-sectional study suggest that use of neuroimaging for dizziness is prevalent across settings. Interventions to optimize the use of neuroimaging must occur early in the patient care journey to discourage guideline-discordant use of CT scans, advocate for judicious MRI use (particularly in ambulatory settings), and account for the effects of price transparency.

History Taking in Non-Acute Vestibular Symptoms: A 4-Step Approach

History taking is crucial in the diagnostic process for vestibular disorders. To facilitate the process, systems such as TiTrATE, SO STONED, and DISCOHAT have been used to describe the different paradigms; together, they address the most important aspects of history taking, viz. time course, triggers, and accompanying symptoms. However, multiple (vestibular) disorders may co-occur in the same patient. This complicates history taking, since the time course, triggers, and accompanying symptoms can vary, depending on the disorder. History taking can, therefore, be improved by addressing the important aspects of each co-occurring vestibular disorder separately. The aim of this document is to describe a 4-step approach for improving history taking in patients with non-acute vestibular symptoms, by guiding the clinician and the patient through the history taking process. It involves a systematic approach that explicitly identifies all co-occurring vestibular disorders in the same patient, and which addresses each of these vestibular disorders separately. The four steps are: (1) describing any attack(s) of vertigo and/or dizziness; (2) describing any chronic vestibular symptoms; (3) screening for functional, psychological, and psychiatric co-morbidity; (4) establishing a comprehensive diagnosis, including all possible co-occurring (vestibular) disorders. In addition, pearls and pitfalls will be discussed separately for each step.

Machine Learning Techniques for Differential Diagnosis of Vertigo and Dizziness: A Review

Vertigo is a sensation of movement that results from disorders of the inner ear balance organs and their central connections, with aetiologies that are often benign and sometimes serious. An individual who develops vertigo can be effectively treated only after a correct diagnosis of the underlying vestibular disorder is reached. Recent advances in artificial intelligence promise novel strategies for the diagnosis and treatment of patients with this common symptom. Human analysts may experience difficulties manually extracting patterns from large clinical datasets. Machine learning techniques can be used to visualize, understand, and classify clinical data to create a computerized, faster, and more accurate evaluation of vertiginous disorders. Practitioners can also use them as a teaching tool to gain knowledge and valuable insights from medical data. This paper provides a review of the literatures from 1999 to 2021 using various feature extraction and machine learning techniques to diagnose vertigo disorders. This paper aims to provide a better understanding of the work done thus far and to provide future directions for research into the use of machine learning in vertigo diagnosis.

Machine learning method intervention: Determine proper screening tests for vestibular disorders

OBJECTIVE

To evaluate the performance of different vestibular indicators in disease classification based on machine learning method.

METHODS

This study use retrospective analysis of the vertigo outpatient database from a tertiary care general hospital. 1491 patients with definite clinical diagnoses were enrolled in this study. Spontaneous nystagmus, head-shaking nystagmus, positional nystagmus, unilateral weakness in caloric test, and gain and saccade in video head impulse test (vHIT) were recorded as variables. Diagnoses were mainly reorganized as acute vestibular syndrome, episodic vestibular syndrome, and chronic vestibular syndrome. The trained random forest model was applied based on exploratory data analysis results.

RESULTS

Random forest accuracies on acute, episodic, and chronic vestibular syndrome are 90%, 81.74%, and 91.3%, respectively. The most important features in acute vestibular syndrome are spontaneous nystagmus, and vHIT variables. In episodic vestibular syndrome, unilateral weakness in caloric test, gain and saccades on lateral semicircular canal are the top three parameters. Lateral vHIT gain, head-shaking nystagmus, and unilateral weakness in caloric test are the main parameters on chronic vestibular syndrome. In acute vestibular syndrome, spontaneous nystagmus and vHIT make major contributions in vestibular disorders distinction. When the disease course prolongation, unilateral weakness and head-shaking nystagmus become increasingly important.

CONCLUSION

Fast clinical test sets including spontaneous nystagmus, head shaking nystagmus, and vHIT should be the first consideration in screening vestibular disorders.

Using Base-ml to Learn Classification of Common Vestibular Disorders on DizzyReg Registry Data

Background: Multivariable analyses (MVA) and machine learning (ML) applied on large datasets may have a high potential to provide clinical decision support in neuro-otology and reveal further avenues for vestibular research. To this end, we build base-ml, a comprehensive MVA/ML software tool, and applied it to three increasingly difficult clinical objectives in differentiation of common vestibular disorders, using data from a large prospective clinical patient registry (DizzyReg).

Methods: Base-ml features a full MVA/ML pipeline for classification of multimodal patient data, comprising tools for data loading and pre-processing; a stringent scheme for nested and stratified cross-validation including hyper-parameter optimization; a set of 11 classifiers, ranging from commonly used algorithms like logistic regression and random forests, to artificial neural network models, including a graph-based deep learning model which we recently proposed; a multi-faceted evaluation of classification metrics; tools from the domain of “Explainable AI” that illustrate the input distribution and a statistical analysis of the most important features identified by multiple classifiers.

Results: In the first clinical task, classification of the bilateral vestibular failure (N = 66) vs. functional dizziness (N = 346) was possible with a classification accuracy ranging up to 92.5% (Random Forest). In the second task, primary functional dizziness (N = 151) vs. secondary functional dizziness (following an organic vestibular syndrome) (N = 204), was classifiable with an accuracy ranging from 56.5 to 64.2% (k-nearest neighbors/logistic regression). The third task compared four episodic disorders, benign paroxysmal positional vertigo (N = 134), vestibular paroxysmia (N = 49), Menière disease (N = 142) and vestibular migraine (N = 215). Classification accuracy ranged between 25.9 and 50.4% (Naïve Bayes/Support Vector Machine). Recent (graph-) deep learning models classified well in all three tasks, but not significantly better than more traditional ML methods. Classifiers reliably identified clinically relevant features as most important toward classification.

Conclusion: The three clinical tasks yielded classification results that correlate with the clinical intuition regarding the difficulty of diagnosis. It is favorable to apply an array of MVA/ML algorithms rather than a single one, to avoid under-estimation of classification accuracy. Base-ml provides a systematic benchmarking of classifiers, with a standardized output of MVA/ML performance on clinical tasks. To alleviate re-implementation efforts, we provide base-ml as an open-source tool for the community.