Vestibular Suppressant Utilization and Subsequent Falls Among Patients 65 Years and Older With Dizziness in the United States

ABSTRACT

Background

Falls pose a significant public health threat to older adults. Due to potential fall risk, guidelines recommend against the routine prescription of several medications commonly used for vestibular suppression, including meclizine and benzodiazepines.

Aims

We aimed to determine the factors associated with vestibular suppressant utilization among patients with dizziness ≥ 65 years of age and subsequent falls.

Methods and Results

A retrospective longitudinal database of US commercial insurance and Medicare beneficiaries was used to study medical claims data (January 1, 2006, through December 31, 2015). Of 190,348 individuals ≥ 65 years old who presented with dizziness, 60,658 (32%) filled a vestibular suppressant prescription (27% anti‐emetics, 73% anxiolytics), of which 20,448 were women (34%) within a month after their dizziness diagnosis. Of those individuals using suppressants, 8% experienced a fall resulting in a medical encounter within 60 days of filling the prescription. After adjusting for sociodemographics and comorbidity, individuals with dizziness who received vestibular suppressants were more likely to experience recorded fall incidents (hazard ratio (HR) 3.33, confidence interval (CI) 1.93–5.72, p < 0.0001), than those who did not receive vestibular suppressants.

Conclusions

Although vestibular suppressants may provide immediate relief from symptoms during an acute vestibular crisis, use is incongruent with guideline‐concordant care for most vestibular diagnoses and is also potentially counterproductive and injurious. Multi‐faceted interventions that engage clinicians and patients are needed to improve the value of care for patients with vestibular disorders to de‐implement the routine use of suppressants for dizziness.

Summary.

• Key points

  • ○
    This study underscores the potential association between guideline discordant use of vestibular suppressant utilization and the likelihood of experiencing subsequent falls.
  • ○
    Of 190,348 individuals ≤ 65 years old who presented with dizziness, 60,658 (32%) filled a vestibular suppressant prescription (27% anti‐emetics, 73% anxiolytics).
  • ○
    Our findings may inform clinical practice and foster shared decision‐making between clinicians and individuals grappling with dizziness.
• Why does this paper matter?

  • ○
    Falls pose a significant public health threat to older adults, leading to widespread disability, severe injuries, and substantial healthcare expenditures.
  • ○
    This study serves as a foundation for future de‐implementation efforts to engage healthcare providers and patients to optimize the use of vestibular suppressants while considering the broader spectrum of safety and care.
  • ○
    We anticipate that this research will foster informed discussions with patients regarding the potential detrimental consequences of vestibular suppressant use, ultimately improving patient care and outcomes.

1. Introduction

Falls pose a significant public health threat to older adults, leading to widespread disability, severe injuries, and substantial healthcare expenditures. These incidents annually impact approximately 4.5 million older residents of the United States and incur significant costs for Medicare, ranging from $15 to $30 billion annually [1, 2]. The risk of injurious falls is heightened among older individuals experiencing dizziness, a common, debilitating, and often difficult to control symptom [3, 4, 5]. Notably, previous studies examining trends in fall‐related injuries among individuals with dizziness have been hampered by the reliance on self‐reported survey data [3], which tends to underestimate the true prevalence of fall‐related injuries.

Medications used to reduce the intensity of symptoms of dizziness, vertigo, and associated motion sensitivity and motion sickness are known as vestibular suppressants. The major drug classes that provide vestibular suppression are antihistamines including meclizine, promethazine, and dimenhydrinate, anticholinergics, and benzodiazepines. These drugs also may have anti‐emetic properties that alleviate autonomic complaints of nausea and vomiting that can accompany vestibular symptoms. Vestibular suppressant medications may provide initial relief for acute dizziness symptoms; however, the American Geriatric Society (AGS) has made recommendations against the routine prescription of several medications commonly used for vestibular suppression, including meclizine, due to its anticholinergic properties, and benzodiazepines, due to risk of dependence, cognitive impairment, falls, and physical injury [6, 7, 8, 9, 10, 11]. Nevertheless, empirical evidence suggests that clinicians in emergency departments and outpatient clinics frequently prescribe vestibular suppressant medications for conditions such as benign paroxysmal positional vertigo (BPPV) and other dizziness symptoms in upwards of 50% of presentations [9, 10, 11]. This disparity between published recommendations and clinical practice remains inadequately documented in the literature, and there is a notable gap in our understanding of the potential impact of such prescriptions on subsequent falls among adults ≥ 65 years old.

The primary objectives of this study were to harness a comprehensive real‐world cohort of individuals over the age of 65 years to explore the factors associated with the utilization of vestibular suppressants following the onset of dizziness and to illuminate the potential association between vestibular suppressant utilization and the likelihood of experiencing subsequent falls within this population. By bridging this critical knowledge gap, we aim to provide valuable insights that can inform clinical practice and enhance the overall well‐being of individuals grappling with dizziness and its associated risks. We anticipate that this research will foster informed discussions with patients regarding the potential detrimental consequences of vestibular suppressant use, ultimately improving patient care and outcomes.

2. Materials and Methods

2.1. Data Source

This study used de‐identified administrative claims data with linked socioeconomic status information from Optum Labs, which includes medical and pharmacy claims and enrollment records for commercial and Medicare Advantage (MA) enrollees. The database contains longitudinal health information on enrollees and patients representing a diverse mixture of ages, ethnicities, and geographical regions across the United States. Commercial beneficiaries in Optum Labs have been shown to be similar in age, race/ethnicity, and sex to the broader US commercially insured population [12]. Since this study involved analysis of pre‐existing, de‐identified data, it was exempted from Institutional Review Board approval at the University of Minnesota.

2.2. Population

The population was comprised of individuals aged 65 years or older who were given a new International Classification of Diseases 9 (ICD9) diagnosis of dizziness and giddiness (780.4) or a vestibular disorder (386.xx) (Table 1) between 2006 and 2015. Diagnosis codes were associated with evaluation and management codes to indicate a provider visit and outpatient provider specialty type was assigned, as previously described [13, 14]. The index date (t ₀) was the date of the initial provider encounter. To select new presentations with consistent follow‐up, individuals had at least 365 days of continuous enrollment before and after t₀ and no dizziness/vestibular diagnoses prior to t ₀. We previously published a description of a large adult dizziness cohort from which older patients for this study were selected [13, 14]. Existing claims data 1 year prior to index date were used to assemble the Charlson Comorbidity Index (CCI) of 19 weighted condition categories [15, 16, 17]. Although care is restricted to that administered in OP and emergency department (ED) settings, we do not have the ability to explicitly exclude nursing home patients who received care in an OP or ED setting.

TABLE 1.

Distribution of vestibular suppressant use over 12 months from time of dizziness presentation by diagnostic category among 65+‐year‐olds (N = 190,348).

	Non‐vestibular suppressant users n = 129,690 (68%)		Vestibular suppressant users n = 60,658 (32%)
Age (years)	n	%	n	%	Crude OR (CI)
65–74	33,071	26	15,874	26	REF
75+	96,619	75	44,784	74	0.96 (0.94–0.99)
Insurance type
Commercial	41,385	32	15,945	26	0.75 (0.74–0.77)
Managed care	88,305	68	44,713	74	REF
Sex
Female	72,756	56	40,210	66	1.54 (1.51–1.57)
Male	56,934	44	20,448	34	REF
Race
Asian	4502	3	1995	3	0.98 (0.92–1.03)
Black	17,055	13	8504	14	1.10 (1.07–1.14)
Hispanic	8558	7	4912	8	1.26 (1.22–1.32)
NH White	82,657	64	37,413	62	REF
Other	16,918	13	7834	13	1.02 (0.99–1.05)
Charlson Comorbidity Index Score
0	32,090	25	12,825	21	REF
1	29,264	23	13,591	22	1.16 (1.13–1.19)
2+	68,336	53	34,242	56	1.26 (1.22–1.28)
Falls within 60 days
No	123,832	98	55,691	92	REF
Yes	3025	2	4967	8	3.68 (3.51–3.85)
Site of first dizziness presentation
ED	29,305	23	18,152	30	1.49 (1.45–1.52)
OP	100,385	77	42,506	70	REF
Therapeutic Class
H6J (anti‐anxiety)			44,499	73
H2F (anti‐emetic)			16,159	27

Abbreviation: NH: Non‐Hispanic.

2.3. Vestibular Suppressant Utilization and Falls Definitions

Vestibular suppressant use was defined as the first occurrence of a prescription within 30 days of a new dizziness diagnosis. See Table 1 for list of National Drug Codes (NDC) of medications selected and used to define vestibular suppressant use based on major therapeutic drug classes of anti‐emetics, including the antihistamine meclizine and motion sickness medications and anti‐anxiety drugs, including benzodiazepines. Cohort characteristics were compared using the crude odds ratios (ORs) for categorical variables. Dizziness diagnoses were categorized into six groups, as defined previously (Supplementary Table S1) [13]. We assessed the distribution of suppressants by diagnosis type. Falls were defined using an established fall identification algorithm comprised of E‐codes (external site of injury codes for accidental falls) plus diagnosis codes for fractures, dislocations, sprains, intracranial injuries, and contusions, as described previously [14]. The timeframe for capturing falls was within 12 months of dizziness diagnosis, with the outcome being the time of the first vestibular suppressant prescription on potential fall risk. All individuals included in the cohort had continuous insurance enrollment for at least 1 year following dizziness diagnosis and were alive for at least 1 year following dizziness diagnosis. We created two fall risk models with individual therapeutic classes (as shown in tables) as well as a model with a combined suppressant use category. Cox proportional hazard modeling was used to assess the time‐dependent factors associated with vestibular suppressants and the factors associated with falls, controlling for patient characteristics. A significance level (α) of 0.05 was used for all analyses. Analyses were completed using SAS 9.4 (SAS Institute, Cary, NC).

3. Results

3.1. Population Characteristics by Vestibular Suppressant

Of 190,348 individuals aged 65 years or older who presented with dizziness, 60,658 (32%) filled a vestibular suppressant prescription within 30 days after their dizziness diagnosis. Of that 60,658, 27% (n = 16,159) obtained an anti‐emetic prescription, and the remainder (73%) obtained an anti‐anxiety medication such as a benzodiazepine. Table 1 presents the characteristics of those who did and did not obtain vestibular suppressants. The factors associated with the receipt of vestibular suppressants by adults with dizziness aged 65 and older included female sex (vs. males), race or ethnicity (individuals identified as Black or Hispanic vs. non‐Hispanic White), higher comorbidity (Charlson Comorbidity Index Score of 2 and 1 vs. 0) and presentation to an emergency department (ED) for first place of dizziness diagnosis (Table 2). Provision of vestibular suppressants by diagnosis type ranged from 30% of those given a symptomatic diagnosis of dizziness and giddiness, not otherwise specified (780.x) to 43% with BPPV and 48% of those diagnosed with vestibular neuritis or labyrinthitis. Further breakdown by diagnosis and by therapeutic class is found in Figure 1.

TABLE 2.

Factors associated with vestibular suppressant use over 12 months from time of dizziness presentation among 65+‐year‐olds.

	OR	95% CI
Age (years) 65–74	REF
75+	0.93	0.91	0.95
Sex
Female	1.58	1.55	1.61
Male	REF
Race
Asian	1.01	0.96	1.07
Black	1.03	1.00	1.06
Hispanic	1.24	1.19	1.29
NH White	REF
Unknown	1.01	0.98	1.04
Charlson Comorbidity Index Score
0	REF
1	1.16	1.13	1.19
2+	1.28	1.25	1.32
Site of first dizziness presentation
ED	1.48	1.44	1.51
OP	REF

Abbreviation: NH: Non‐Hispanic.

FIGURE 1.

Distribution of vestibular suppressant use within 1 month from time of dizziness presentation by diagnostic category among 65+‐year‐olds (N = 190,348).

3.2. Factors Associated With Falling

When adjusting for all other factors (Table 3), individuals with dizziness who filled a vestibular suppressant prescription of either therapeutic class were more likely to have a recorded fall encounter than those who did not fill a vestibular suppressant prescription (hazard ratio (HR) 2.17, confidence interval (CI) 1.26–3.74 for anti‐emetic (H6J) vs. no suppressants and HR 4.13, CI 2.40–7.12 for anti‐anxiety (H2F) vs. no suppressants). Other factors associated with falls in the adjusted Cox model included higher comorbidity score and ED visit as the site of first dizziness presentation vs. OP (HR 0.50, CI (0.47–0.54)). Individuals identified as Hispanic, Asian, or Black were significantly less likely than non‐Hispanic white individuals to have a recorded fall encounter after a vestibular suppressant prescription (Table 3).

TABLE 3.

Cox model of factors associated with falling among 65+‐year‐olds with dizziness diagnoses.

	HR	95% CI		p
Age (Years)
65–74		REF
75+	1.04	(0.97–1.15)		0.20
Sex
Female	0.95	(0.89–1.03)		0.06
Male		REF
Race
Asian	0.55	(0.49–0.67)		< 0.0001
Black	0.91	(0.84–0.98)		0.02
Hispanic	0.76	(0.67–0.85)		< 0.0001
NH White		REF
Unknown	1.01	(0.92–1.09)		0.97
Charlson Comorbidity Index Score
0		REF
1	1.43	(1.30–1.58)		< 0.0001
2+	1.49	(1.37–1.62)		< 0.0001
Site of first dizziness presentation
ED		REF
OP	0.50	(0.47–0.54)		< 0.0001
Therapeutic Class a
No suppressant use		REF
H6J (anti‐anxiety)	4.13	(2.40–7.12)		< 0.0001
H2F (anti‐emetic)	2.17	(1.26–3.74)		0.01

^a Overall hazard ratio (HR) 3.33, confidence interval (CI) 1.93–5.72, p < 0.0001 in comparison to their counterparts with dizziness who did not receive such medications. Bold values significant at ≤ 0.05.

4. Discussion

Our study presents a novel exploration of vestibular suppressant use patterns among adults with dizziness aged 65 years and older in the United States, offering valuable insights into the association between these medications and subsequent falls among patients with dizziness. We found that a substantial proportion (32%) of individuals aged 65 and older who sought medical attention for dizziness filled a vestibular suppressant prescription within a month of their diagnosis. Of those prescribed vestibular suppressants, 26% were provided with an anti‐emetic such as meclizine or dimenhydrinate, while the remaining 74% were prescribed an anxiolytic such as a benzodiazepine. We found that 8% of those individuals who received a vestibular suppressant experienced a fall resulting in a medical encounter compared to 2% who did not receive a vestibular suppressant. After controlling for sociodemographics, comorbidity, and site of presentation, individuals with dizziness who filled vestibular suppressant prescriptions remained more likely than those who did not experience recorded fall incidents (hazard ratio (HR) 2.17, confidence interval (CI) 1.26–3.74 for anti‐emetic (H6J) vs. no suppressants and HR 4.13, CI 2.40–7.12 for anti‐anxiety (H2F) vs. no suppressants). While both drug classes increased fall risk, the hazard ratio was higher for benzodiazepines than anti‐emetics.

Most falls result from an interaction between individual characteristics, such as higher comorbidity, that increases an individual's propensity to fall, and acute mediating risk factors that provide the opportunity to fall [15, 16]. Increased fall risk has previously been reported among community‐dwelling older women compared to men [16]. Dizziness is a well‐documented risk factor for self‐reported falls and falls with injury [17]. Adults aged 60 years or older who fall because of dizziness or loss of balance were observed to be more than twice as likely to be hospitalized than if the fall occurred after slipping or tripping [5]. Others have also found that patient factors including acute and chronic illnesses, abnormalities of gait or balance, and medications can increase the risk of falls and hip fractures in older adults, especially in tertiary care academic medical centers, within 30 days after admission [13]. Our findings in this study illustrate an additional longer‐term association between the use of vestibular suppressants and subsequent falls after either outpatient or emergency department dizziness presentations. Although the medications are typically intended to alleviate symptoms and patient suffering, individuals who received vestibular suppressants were 3–4 times more likely to experience recorded fall incidents in comparison to their counterparts who did not receive such medications. Consistent with observations in general populations, factors such as higher comorbidity scores, age of 74 years or older, female sex, and white race were also identified as contributing to an elevated risk of falls in our cohort [17, 18, 19, 20].

Despite the availability of established guidelines like the AGS Beers Criteria for Potentially Inappropriate Medication (PIM) Use in Older Adults, which have been widely approved by clinicians, educators, researchers, healthcare administrators, and regulators [5, 11, 18], the use of vestibular suppressants remains prevalent by healthcare providers in both ED and outpatient settings. These criteria, initially implemented over a decade ago in 2011 and updated on a three‐year cycle, aim to guide medication choices for adults aged 65 and older across various care settings, excluding hospice and palliative care. AGS cites moderate evidence in making a strong recommendation against use of meclizine and benzodiazepines among older adults. The antihistamine meclizine is highly anticholinergic and cumulative exposure to anticholinergics is associated with risk of falls, delirium, and dementia [21, 22]. Older adults have increased sensitivity to and decreased metabolism of benzodiazepines, and these drugs also carry risk of physical dependence, cognitive impairment, falls, and injuries [21, 22]. Despite these clear guidelines, frequent provision of vestibular suppressants to older adults persists in clinical practice [23]. We observed a higher (23.3%) prevalence of use of anxiolytics, comprised predominantly of benzodiazepines, in our cohort of dizziness patients, which even more strongly suggests a need for targeted interventions to align prescription practices with evidence‐based recommendations. Additionally, we found that anxiolytics were more strongly associated with the hazard of falling than anti‐emetics (HR 4.13, CI 2.40–7.12 for H2F (anti‐anxiety) vs. no suppressants; HR 2.17, confidence interval (CI) 1.26–3.74 for H6J (anti‐emetic) vs. no suppressants). Additional factors that emerged as being associated with the receipt of vestibular suppressants among older adults with dizziness in our cohort included female sex, higher comorbidity, Black race, Hispanic ethnicity, and dizziness evaluation in the outpatient setting. These factors will be pertinent when designing efforts to decrease inappropriate use.

As there are instances when symptomatic treatment with vestibular suppressants is helpful, we likely observed a mix of appropriate and inappropriate prescribing in our cohort. In the first several days of an acute vestibular crisis, as results from vestibular neuritis or labyrinthitis, patients can experience debilitating spontaneous vertigo that improves as vestibular compensation occurs. Other disorders, such as Meneire's disease, result in hours‐long vertigo attacks. In such instances, or before a diagnosis is reached, it may be appropriate to provide a short course of medication for relief of intense symptoms. However, longer‐term use of vestibular suppressants delays or prevents vestibular compensation and may promote symptom chronicity [24]. Recently, a meta‐analysis determined that, for relief of acute vertigo, single‐dose meclizine is more effective than benzodiazepines; further, benzodiazepines are not associated with an improvement in outcomes for acute vertigo over placebo [25]. These findings argue for the de‐implementation of benzodiazepines for acute vertigo [25]. Further, clinical practice guidelines specifically recommend against the use of vestibular suppressants for the most common form of peripheral vertigo, BPPV [22]. Nevertheless, among individuals with a diagnosis of BPPV, 43% received a vestibular suppressant, emphasizing that use is not restricted to indicated diagnoses (Figure 1). Rather, patients should be offered canalith repositioning procedures by the evaluating clinician or by referral to Physical Therapy (PT) or another capable provider [26, 27]. Alternatively, patients may be able to receive information about these procedures through patient education materials at the time of discharge from the emergency department or primary care clinic even if specialty clinic or physical therapist is not available.

Both implementation and de‐implementation efforts are essential in managing care for patients with dizziness. Implementation efforts include achieving an accurate diagnosis and initiating the correct, etiologically based therapy to address the underlying disease, which is ultimately more patient‐centered than managing symptoms with vestibular suppressants [28, 29]. Although the AGS Beer's criteria underscore the risks of vestibular suppression among individuals over age 65, other interventions may also be effective. Since we observed that vestibular suppressants are associated with fall risk for individuals with dizziness, de‐implementation efforts are also warranted. De‐implementation focuses on reducing low‐value care, especially treatments that are harmful or lack evidence of effectiveness.

De‐implementation strategies may be most effective when employing multi‐component interventions that engage both clinicians and patients in reducing overuse [26]. Clinicians may benefit from education, feedback, and clinical decision support [26, 30]. For instance, disseminating diagnosis‐based clinical practice guidelines, such as those for BPPV, has been shown to reduce inappropriate prescribing. Following the 2008 BPPV guideline publication [27], research by Dunlap et al. noted a drop in antivertigo or anti‐emetic prescriptions during adult ambulatory visits from 50% in 2007–2009 to 16% in 2013–2015 [22]. However, many patients initially presenting with dizziness receive non‐specific diagnoses, and no current guideline exists for managing acute vertigo, which leaves clinicians without a clear course of action. Offering alternatives to medication, like vestibular physical therapy, could be beneficial [14]. In the US, the AGS has joined the Choosing Wisely campaign, advising against benzodiazepine use in older adults for insomnia, agitation, or delirium [31]. However, attempts to implement this guidance through behavioral economics approaches in US clinics have highlighted the challenges of reducing low‐value medications in busy clinical settings. Comprehensive and nuanced efforts will be necessary to effectively address this issue.

Despite the compelling findings, this study is not without limitations. The claims‐based nature of the research focused on dizziness and falls resulting in medical encounters, potentially selecting patients with more severe symptoms or injuries. E‐codes are frequently missing in claims data and reliance on E‐codes alone to indicate fall severity is insufficiently sensitive [14]. While we employed an established method that captures injuries attributable to falls, we cannot definitively attribute all injuries to falls, potentially leading to an overestimation of fall prevalence. However, falls not captured in claims data may translate to an underestimation of falls. In order to maintain consistent billing and dizziness diagnoses codes throughout the course of this study we restricted the date range of the data from 2006 to 2015. However, we believe our results remain relevant for several reasons. The medications of focus—benzodiazepines and anti‐emetics—are still commonly prescribed, and their potential to contribute to falls among older adults remains a critical issue. Furthermore, risks associated with these medications, particularly in vulnerable populations, have not significantly changed. Additional limitations include unmeasured confounding, such as by prior falls, or by indication, as individuals who received medication may have had more severe dizziness, although use may reflect complex practice patterns that differ between clinicians. Our study concentrated on filled prescriptions of vestibular suppressants, and we could not ascertain medication adherence rates or over‐the‐counter use. We were also unable to capture the chronic use of suppressants prior to dizziness diagnosis. Although others have focused on polypharmacy rates, which would undoubtedly compound fall risk and outcomes depending on the mix of medications [32, 33], future research should delve deeper into the pharmacological aspects that might influence fall risk among vestibular suppressant users.

5. Conclusion

In conclusion, this study provides valuable insights into the patterns of vestibular suppressant use among older adults with dizziness and their potential impact on falls. Our findings emphasize an association between these medications and falls above that ascribed to dizziness alone, but further research is needed to understand the complex interplay between medication use and fall risk comprehensively. This study serves as a foundation for future de‐implementation efforts to engage healthcare providers and patients to optimize the use of vestibular suppressants while considering the broader spectrum of safety and care.

Author Contributions

Drs. Adams and Marmor had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: All authors. Acquisition, analysis, or interpretation of data: Adams, Marmor. Drafting of the manuscript: Adams, Marmor. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Karaca‐Mandic, Marmor. Obtained funding: Adams. Supervision: Adams and Marmor.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Table S1. Codes used for Classification of Dizziness Diagnoses and Therapeutic Drug Class by National Drug Codes (NDC) and Falls.