Economic burden of vertigo: a systematic review

Background Vertigo, a highly prevalent disease, imposes a rising burden on the health care system, exacerbated by the ageing of the population; and further, contributes to a wide span of indirect burden due to reduced capacity to work or need of assistance in activities of daily living. The aim of this review was to summarise the evidence on the economic burden of vertigo. Methods All original studies published between 2008 and 2018 about the economic evaluation of peripheral or central vestibular vertigo in developed countries were considered eligible, unrestricted to setting, health care provider, or study type. Results The electronic search in three databases identified 154 studies from which 16 qualified for inclusion. All studies presented partial economic evaluation referring to a variety of vestibular vertigo including unspecified vertigo. Nine studies presented monetised cost results and seven studies reported health care utilization. Direct costs derived mainly from repeated and not well-targeted consultations at all levels of health care, excessive use of diagnostic imaging, and/or of emergency care. Considerable societal burden was caused by decreased productivity, mainly due to work absenteeism. Conclusion To the best of our knowledge, this is the first systematic review of the existing evidence of the economic burden of vertigo. The lack of conclusive evidence revealed apparent targets of future research. First, studies of diagnostics and therapies for vestibular disease should include cost-effectiveness considerations. Population-based studies of health services utilization should include simple vestibular assessments to get more reliable estimates of the burden of disease and associated costs on the level of the general population. Further, clinical and population-based registries that include patients with vestibular disease, should consider collecting long-term data of societal burden. Primary data collection should increasingly include assessment of health care utilization e.g. by linking their diagnoses and outcomes to routine data from health insurances.


Background
Vertigo and dizziness, belonging to the most frequent symptoms with an estimated lifetime prevalence of 17-30% [1], cover diseases and conditions of various origin [2,3]. In a narrower sense, vertigo refers to peripheral or central vestibular diseases with a lifetime prevalence of up to 10% [1,4] and a yearly incidence of 1.4% [4]. The most prevalent types of peripheral vestibular vertigo are benign paroxysmal positional vertigo (BPPV), Meniere's disease (MD), vestibular neuritis and bilateral vestibulopathy; vestibular migraine is one of the most common examples of central vestibular vertigo [3,5].
Arguably, vertigo and dizziness are among the main drivers for health care utilization from primary care [6][7][8] through specialist care [9] to tertiary level hospitals [10]; therefore may have a high impact on direct costs in industrialized countries. The increasing prevalence of vertigo in older population [11] further contributes to this burden of health care [12].
Individuals with acute symptoms may present at all levels of the health system including emergency services [13,14]; however most instances of vertigo might be diagnosed and treated at the primary care level [15]. This is often not the case, leading to unjustified diagnostic procedures, prolonged time to diagnosis and repeated specialist consultations [6]. One major driver of direct health care costs may be the overutilization of imaging procedures, which actually would have a well-defined but limited role in differentiating vestibular disease from rare but life-threatening conditions such as stroke [16]. Regarding therapy in BPPV, the most prevalent type of vertigo in older adults [4], liberatory manoeuvres [17,18] may bring fast relief [19]. Pharmacotherapy has its place in a limited number of pathologies [20]; vestibular rehabilitation [21] should be offered to all patients with vestibular deficiency [22].
Regarding indirect costs, vertigo can be a reason for sick leave and occupational disability. In a study from the United Kingdom (UK) and Italy, patients reported a mean of 7 days absence from work due to dizziness in the previous 6 months [23]. From those patients still working, over 50% felt that their work efficiency had dropped [23]; over one fourth of them had changed their jobs and 21% had quit work [23]. Likewise, in Belgium, more than half of the patients of a tertiary dizziness center reported having been on sick leave due to dizziness, and 12% were completely unable to return to work [24].
Also, vestibular disease may cause considerable restrictions of activities of daily living [23] and quality of life [25]. Loss of quality of life from vestibular disease was recently estimated to amount to a total of 64,929 USD lifetime economic burden per patient, or in a total lifetime societal burden of 227 billion USD for the USA population over 60 years of age [26].
As financial burden of disease seems to be considerable for vertigo and dizziness, we aimed to summarise the evidence from available quantitative studies in a systematic way. Specifically, we aimed to summarize information on costs arising from diagnosis, referral or therapy.

Methods
In this review the Preferred Reporting Items for Systematic Reviews and Meta-Analyses [27] was followed.

Search strategy and eligibility criteria
All original studies regarding any kind of economic evaluation of vertigo were considered irrespective of study design or perspective (e.g. payer or society).
The detailed search strategy for the electronic databases Medline, EMBASE, and Cochrane Library is presented in the Additional file 1 Beside the general terms "vertigo" or "dizziness", the search terms were compiled to cover the above listed most prevalent vertigo types of peripherial or central vestibular origin, and functional vertigo; including both their referring Medical Subject Headings (MeSH) terms and additional free text. Restricting the search to title and abstract ensured to focus on studies handling vertigo or dizziness as the main topic of the study. Vestibular Schwannoma was excluded with respect to its surgical treatment being not comparable with the conservative management of all other vertigo types. MeSH terms representing a broad range of economic evaluation were completed with a free text search.
With respect to the comparability of economic results, developed countries according to the categorization of the United Nations Statistics Division [28] were selected. The hits were restricted to being published in English in the last 10 years, i.e. between October 2008-01. October 2018.
Full texts were retrieved via the online library service of the Ludwig-Maximilians-Universität München. We contacted two authors for further information unsuccessfully. The search in the electronic databases was not extended by a handpicked search. Study selection was performed by two independent researchers (XW and EK). Disagreement was resolved by discussion or in lack of consensus by the decision of a third researcher (EG).

Data collection
A Microsoft Excel form was prepared (XW) for collecting data of study characteristics, vertigo type(s), and outcome indicators. Assessing methodological quality and risk of bias, the list of Consensus on Health Economic Criteria (CHEC) [29] was applied. One researcher (XW) performed the data extraction and a second researcher (EK) checked the data.

Conversion of the economic results
Costs expressed in national currencies and in different price years were reported both in original form and converted into 2016 USD using The Campbell and Cochrane Economics Methods Group Evidence for Policy and Practice Information Coordinating Centre (CCEMG -EPPI-Centre) Cost Converter [30]. The converter adjusts the price year according to the Gross Domestic Product deflator index; while conversion between countries/currencies is based on Purchasing Power Parities, in accordance with the recommendation of the Cochrane Handbook for Systematic Reviews [31].
In case of studies reporting health care utilization in units as an outcome, health care utilization in units as an outcome, despite considerable efforts we were not able to retrieve historical unit prices for converting reported units into monetary estimates.

Study selection
The electronic search identified 154 studies from three databases. After removing duplicates, 104 studies were screened by title and abstract according to the eligibility criteria. Inclusion was discussed in 21 cases, resulting in 35 studies qualifying for full-text reading. Disagreement was then solved by discussion in 17 cases, and in lack of consensus in six cases, the third researcher made the decision. Sixteen studies were included in the review. Figure 1 presents the flow of study selection.
The studies reported the burden of vertigo by variuos means: resource use per patient, aggregated to or in rate  of a certain population; monetised cost either measured or estimated per patient, per diagnostic investigation and/or per positive yield, or on different agggregation level. The applied population covered a wide range from general population, nationally representative patient sample or overall national health care data, or patient population diverse in terms of the above detailed diagnoses and health care services (Table 1).
With two exceptions [41,47], the studies were observational; quality criteria items referring to an intervention being not applicable. This resulted in a median quality score of six from 19 total of the CHEC (Additional file 2). One of the studies [44] conducted a sensitivity analysis; one study [42] applied inflation adjustment.

Medical consultations
From the unselected adult (17-79 year-old) population 1.8% had a consultation in the previous 12 months [40]. In Germany these consultations were roughly equally distributed between primary health care and specialists [40]; in the USA the contribution of primary health care providers was 14.3% [38].
Separated according to the level of health care, the range of primary care consultations was 1.1-2.6 occasion, and the range of specialist consultation was 0.8-1.8 [35] within the 3 month observation period in European countries. More than two consultations occurred in 61.3% of the vertigo patients [37]. Vestibular vertigo patients had more consultations than patients with nonvestibular vertigo [40]. Co-morbid anxiety increased the number of consultations up to 6.6 (SD 5.4) within the previous 12 months [46]. Higher age also contributed to more consultations [35,40].

Emergency care
Occasions of emergency care visits ranged from 0.1 to 0.6 [35] within 3 months in European countries, or 0.1 (SD 0.5) to 0.4 (SD 0.8) within 12 months [44] in the USA, the latter corresponding to mean costs of 94-274 (range 0-2374) USD per case [44]. In a USA national overview, 3.9 million emergency care visits in 2011 resulted in 3.9 billion USD total costs, i.e. on average 1004 USD per patient and visit [42].

Hospitalisation
Hospitalisation in Germany due to vertigo was reported by 1.9% of the unselected adult population [40]. Lifetime hospitalisation occurred in 10% of vestibular vertigo patients and in 5% of patients with non-vestibular vertigo [40].
The number of hospital days ranged from 0.4 to 1.7 days during a 3 month observation period [35] in European countries. In the USA, BVD patients had 1.4 (SD 0.8) hospital visits within 12 months, causing a 203 (range 0-348) USD cost [44]. Advanced age contributed to longer hospital stay [35].

Diagnostic procedures
Patients in Germany underwent on average 3.2 (range 0-6) instrumental diagnostic procedures [37], the older age group receiving significantly less [32,37]. Economic reporting about diagnostic procedures focused on imaging, being the most frequently used [37], and the most expensive diagnostic procedures accounted for 12% of the total annual costs for dizziness visits [42]. Prior to referral to a German tertiary balance centre, 82% of the patients received either magnetic resonance imaging (MRI) or computed tomography (CT) [37].
CT was performed in 22.8% [42] and up to 50% [39] of the vertigo patients, depending on the setting, i.e. general patient population or emergency patients, respectively. Each 10-years step in age increased the probability of scanning to 1.4 times [33]. CT use showed an increasing trend in a national-level USA study [42], with an almost two-thirds increase between 2005 and 2009. However, the yielding rate of CT for vestibular diseases was reported to be low, from 0.74% [33] to 3.6% [34]. This resulted in a high diagnostic cost for positive findings, ranging from 54, 540 USD [34] to 176,720 USD [33].
Other diagnostic methods were less frequent than imaging: head impulse test was performed in 5% of the patients and the Dix-Hallpike manoeuvre in 31.4% [39]. Among the listed diagnostic methods, head impulse test had the highest yielding rate for vestibular disease of 29% [39].

Therapy and medication
Vertigo patients had on average 1.8 (range 0-8) different therapies; the most frequent therapy was medication (61.0% of all patients), with mean of 1.8 (range 0-17) drugs per patient [37]. Physical therapy was prescribed in 41.3% of the patients [37]. The liberatory Epley manoeuvre was performed in 15.3% of BPPV patients [39].

Indirect costs of vertigo
Loss of working days (Table 3) ranged from 13.1 (SD 14.6) days during a 3 month observation period [35], to 69 (SD 106) days in 12 months [44], the latter corresponding to a mean productivity loss of 12,542 USD [44]. Sick leave was significantly higher in vestibular vertigo patients than in patients with non-vestibular vertigo [40]. Also, 69.8% of the patients reported they had to reduce their workload, 4.6% changed their job and 5.7% quit work [35]. In Norway, almost 1 % of the overall long-term sickness absence was caused by vertigo [43]. Among the affected patients, 23% of women and 24% of men obtained a disability pension [43]. Regarding non-monetary burden of disease, both vestibular vertigo patients and patients with non-vestibular vertigo reported avoiding leaving the house (18.5% and 10.1%, respectively), experienced an interruption of daily activities (40.3% and 11.5%, respectively), and perceived reduced quality of life [40] and relevant restrictions in daily life [44]. Concomitant anxiety was reported to further impair social life [46]. Patients with chronic vestibular loss reported an increased frequency of falls [44].
Two studies report on both direct and indirect cost, indicating that the latter may be even higher [44,45].

Total costs
Two studies provided a wider view of the total cost of vertigo. A UK study [45] estimated as comprehensive cost of MD 3341-3757 GBP [4865 -5470] per patients annually. A USA study included emergency visits and hospitalisation plus an estimation of lost working days, adding an average annual economic burden of 13,717 USD per BVD patients [44].

Discussion
This review is the first to explore the economic burden of vertigo and dizziness including direct and indirect costs. We found large heterogeneity in respect to the investigated sector of the health care system, the type of the vertigo, and the cost components; the quality of studies was mostly mediocre. Annually up to 9.6 visits were reported at the primary care provider [35], up to 7.2 visits at the specialist [35], with up to 6 instrumental diagnostic procedures [37]. Imaging was performed in up to 82% of the patients [37] with a cost of up to 164, 700 [176,720] USD per positive yield [33]. Up to 2.4 presentations occurred at the emergency department [35], leading to up to 6.8 hospital days [35]. The trend of the number of ED visits, and imaging due to the vertigo was rising [38,42]. Studies identified an increase by aging of the population in vertigo prevalence [32,36,38], in vertigo-related health care demand [35,38,40,42] and in imaging performed [33]; however, not in the number of other investigations [37].
The original studies reported three main drivers of increased direct costs: unnecessarily repeated consultations of primary, and specialist care, and referrals, where primary care would have been sufficient; excessive use of diagnostic imaging, and excessive use of emergency care, the latter mainly in countries where statutory affordable health insurance is not broadly available. The number of referrals and consultations largely depends on the respective health care system. Typically, a system of statutory health insurance will regulate reconsultation and referral, however, in rather liberal systems such as Germany, referral is easy and sometimes more cost effective for the primary care physician [48]. Also, in an earlier study we found that primary care providers do not always feel competent to diagnose and treat vertigo [49]. Consequently, patients with vertigo may undergo several consultations and referrals without clear diagnosis or therapy [6,35,40].
Second, several studies examined the excess costs of extensive imaging procedures [33,34,42]. It is reasonable to assume that physicians do not want to overlook life-threatening diseases. In absence of defined clinical pathways and in systems where imaging is broadly available, imaging seems like an easy solution. In contrast, low-cost examination techniques that require a certain skill set, such as the head impulse test [50] seem to be underutilized [39].
Third, consulting emergency care is not unusual because vertigo is a worrying symptom. However, overuse of the emergency department because of vertiginous symptoms may be a direct result of a system where primary care is not always affordable or available [51].
The limited data on indirect costs due to vertigo indicate that it may be considerable [35,44,45].
Considering the three main drivers of direct costs, there is no easy solution to the problem. Original studies recognising the problem of unnecessary or not welltargeted investigations argued for guidelines or managed care [33,34,42]. Application of clinical practice guidelines may provide benefit even in terms of preventing unnecessary diagnostic investigations, and supporting evidence-based resource use; thus, contributing to savings [52,53]. The German Association for Primary Care has put forward a set of guidelines for management of vertigo in the primary care setting [54]. Likewise, guidelines exist in the USA [55], in the Netherlands [56], in Spain [57], in Croatia [58], or in China [59]. Treating vertigo frequently needs a multidisciplinary approach with combined expertise from several medical and therapeutic professions. Capacity building may reduce the vertigo-related direct costs through promoting evidencebased practices and increasing the knowledge base. However, due to the fragmentation of health care, PCPs often lack resources and time to coordinate such a team-based approach [60]. There is therefore an urgent need to focus economic research on the detected fields of main cost drivers, and investigate the effectiveness of care pathways and managed care approaches.

Limitation
The selected studies targeted various types of vestibular diseases and various cost components, providing a fragmented and incomplete picture. With respect to the limited number of cost reporting, we included studies reporting health care utilisation. Due to difficulties of finding historical unit price information, these were not converted to a monetised form. Though technical and equipment costs might be similar e.g. for imaging, other background aspects of national regulation and financing of the health care system may divert these prices in a broad range [61]. Thus, the financial burden may widely deviate in different countries even by similar prevalence of vertigo and similar frequency of health care utilisation. This limits the value of these data regarding the economic burden of vertigo. The gatekeeping function of primary care and the direct accessibility of specialist consultation is regulated differently depending on the country, thus, the utilisation of consultations on different levels of health care are not directly comparable.

Conclusion
To the best of our knowledge, this is the first systematic review of the existing evidence of the economic burden of vertigo. Results demonstrated that vertigo contributes to an increasing trend of direct health care costs, due to the ageing of the population. Three main cost drivers due to insufficient diagnostic skills were identified. Repeated and not well-targeted health care consultations on all levels, excessive use of expensive diagnostic imaging, and unnecessary assignment to emergency care. Several studies demonstrated that the ageing of the population contributes to an increasing trend of direct health care costs in persons with vertigo. The main result, i.e. that there is no conclusive evidence expressing an overall economic burden, may seem disappointing at first. It is, however, the objective of any systematic review not only to summarize existing knowledge but also to summarize apparent gaps of evidence. There are several direct consequences from this gap: 1. Future studies of new or established diagnostic devices and therapies for vestibular disease should include cost-effectiveness considerations. 2. Population-based studies that focus on health care utilization should include simple assessments of vestibular function to get more reliable estimates of burden of disease and associated costs on the level of the general population. 3. Clinical and population-based registries that include patients with vestibular disease, should invest more effort in the exact characterization of disease and consider collecting long-term patient-reported outcomes, absence from work and other types of societal burden. 4. Primary data collection should increasingly include assessment of health care utilization e.g. by linking their diagnoses and outcomes to routine data from health insurances.
Additional file 1: Detailed search strategy.
Additional file 2. Assessment of study quality and risk of bias by the Consensus on Health Economic Criteria*.