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

Key Points

Question

How does the use of neuroimaging differ for patients with dizziness who present to outpatient clinics vs those who present to emergency departments (EDs)?

Findings

In this cross-sectional study of 805 454 individuals evaluated for dizziness, the neuroimaging rates were 35% after ED presentations and 15% after outpatient presentations; use of neuroimaging was significantly associated with age, race and ethnicity, and comorbidity. Use of a computed tomography (CT) scan of the head was prevalent across settings, whereas magnetic resonance imaging (MRI) was more often performed after outpatient presentations, accounting for 70% of total neuroimaging spending.

Meaning

Although the optimization of neuroimaging for dizziness has focused on the overuse of head CT scans in the ED, interventions are needed to promote guideline-concordant MRI and CT use in outpatient settings, where most dizziness presentations occur.

Abstract

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.

This cross-sectional study uses commercial and Medicare Advantage claims to characterize neuroimaging use, timing, and spending as well as the factors associated with imaging acquisition within 6 months of presentation for dizziness in outpatient clinics vs emergency departments in the US.

Introduction

With technologies being widely available, use of neuroimaging for the highly prevalent symptom of dizziness^1,2,3 has increased considerably in recent decades. From 1995 to 2011, as annual emergency department (ED) visits for dizziness increased 97% to 3.9 million, the proportion of these ED visits using imaging technology increased from 10% to nearly 40%, with head computed tomography (CT) scans being the predominant modality.¹ Attention has been drawn to the potential overuse, unwarranted variations, and costs of neuroimaging for patients with dizziness in ED settings.^1,2 The optimal clinical pathways for an accurate and efficient diagnosis of dizziness remain undefined, but clinicians and patients generally use neuroimaging out of concern for missing underlying intracranial and cerebrovascular lesions. Our group and others have observed that most individuals with dizziness first present to outpatient clinics, where patterns of neuroimaging use are less characterized.^3,4

Because high use of health services with potentially limited benefit increases costs and lowers the value of US health care,⁵ there are several reasons to reconsider the patterns of neuroimaging use for dizziness. Neuroimaging often fails to reveal the cause of dizziness. Prevalent underlying conditions (including medical and peripheral vestibular disorders) lack abnormalities that might be detected using neuroimaging, and central lesions (including cerebellar or brainstem strokes, vertebrobasilar insufficiency, demyelination, and posterior fossa neoplasms) are relatively uncommon.^6,7 Although technically noninvasive, neuroimaging is costly and not without potential physiological risks (radiation) or psychological risks (incidental findings).^8,9 Diagnostic accuracy also varies by neuroimaging modality and timing. A head CT scan provides false reassurance during evaluations for dizziness given its low sensitivity for posterior fossa infarcts (16%) compared with diffusion-weighted magnetic resonance imaging (MRI) (83%).^10,11,12 Magnetic resonance imaging may also be falsely negative for small infarcts within 48 hours of onset.¹³ Accordingly, radiologic guidelines recommend appropriately timed MRI vs CT scan for evaluating the central causes of dizziness or vertigo, and temporal bone CT scan vs head CT scan for labyrinthine dehiscence or erosion.^14,15

Best practices and policies need to be informed by clinical use and spending data throughout the patient care journey. This study analyzes the use of neuroimaging across episodes of care for dizziness among a broad population of US commercial insurance and Medicare Advantage (MA) enrollees. The objectives were to characterize neuroimaging use, timing, and direct spending as well as the factors associated with acquisition within 6 months of presentation for dizziness to outpatient vs ED settings. Findings are placed in the context of health care policy initiatives to optimize resource use and deimplement ineffective care.^16,17

Methods

Data Source

This study used deidentified administrative claims data with linked socioeconomic status information from the OptumLabs Data Warehouse, which includes medical and pharmacy claims, laboratory test results, and enrollment records for commercial insurance and MA enrollees. The database contains longitudinal health information on enrollees and patients, representing a diverse mixture of ages, races and ethnicities, and geographical regions across the US.¹⁸ Because this study involved analysis of preexisting, deidentified data, it was exempted from institutional review board approval at the University of Minnesota. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Population

The population comprised individuals aged 18 years or older who were given a new International Classification of Diseases, Ninth Revision diagnosis of dizziness and giddiness (code 780.4) or a vestibular disorder (code 386.xx) between January 1, 2006, and December 31, 2015. Diagnosis codes were associated with evaluation and management codes to indicate a clinician visit. The index date was the date of the initial clinician encounter. To select new presentations with consistent follow-up, individuals had to have had at least 365 days of continuous enrollment before and after the index date and no dizziness or vestibular diagnoses prior to the index date. A description of this cohort has been published previously.³

Neuroimaging and Cost Definitions

Neuroimaging was defined as the occurrence of a CT scan, CT angiography, MRI, or magnetic resonance angiography of the head or neck or the occurrence of cerebrovascular ultrasonography (USN) from first dizziness diagnosis to 6 months after the index date (eTable in the Supplement). As the first study, to our knowledge, measuring the use of neuroimaging for dizziness longitudinally across settings, we selected a 6-month duration for the episode of care^19,20,21,22 based on sensitivity analyses estimating that more than 90% of neuroimaging was performed within the first 6 months of a 12-month window. We assessed overall neuroimaging use and the total number of neuroimaging tests. To analyze spending trends, we constructed distinct time periods of neuroimaging use at the index date, at 6 months, and overall. We computed a total cost that included patient out-of-pocket spending and total health plan costs.

Statistical Analysis

Statistical analysis was conducted from from October 1, 2020, to September 30, 2021. Cohort characteristics were compared using the χ² test for categorical variables. An unadjusted Kaplan-Meier analysis was used to compare time to imaging events. Multivariable regression was used to assess the factors associated with neuroimaging. Cox proportional hazards regression modeling was used to evaluate the association of outcomes with time to neuroimaging event in outpatient settings, controlling for patient factors and clinician specialty. Analyses were completed using SAS, version 9.3 (SAS Institute Inc).

Results

Population and Neuroimaging Use

We observed 805 454 individuals (502 055 women [62%]; median age, 52 years [range, 18-87 years]) with a diagnosis of dizziness (Table 1; eFigure in the Supplement), as previously described.³ Within 6 months of presentation, 156 969 patients (20%) underwent neuroimaging, 115 974 (74%) of whom received 1 neuroimaging test and 40 995 (26%) of whom received 2 or more neuroimaging tests. The 6-month neuroimaging rate for ED presentations (35% [65 738 of 185 338]) was 21% higher (95% CI around the difference, 20.5%-20.9%) than outpatient presentations (15% [91 231 of 620 116]). The median time from the index date to neuroimaging was 0 days (95% CI, 0-2 days) for ED presentations and 10 days (95% CI, 9-10 days) for outpatient presentations (Figure).

Table 1. Characteristics of Patients With a Diagnosis of Dizziness Who Did or Did Not Undergo Neuroimaging Within 6 Months of Presentation.

Characteristic	Patients, No. (%)
	Total population (N = 805 454)	Imaging
		No (n = 648 485)a	Yes (n = 156 969)a
Age, y
18-39	180 795 (22)	156 848 (87)	23 947 (13)
40-49	157 038 (20)	129 681 (83)	27 357 (17)
50-59	175 021 (22)	139 571 (80)	35 450 (20)
60-64	72 111 (9)	56 088 (78)	16 023 (22)
65-74	57 360 (7)	45 366 (79)	11 994 (21)
≥75	163 129 (20)	120 931 (74)	42 198 (26)
Insurance type
Commercial	634 707 (79)	516 390 (81)	118 317 (19)
Medicare Advantage	170 747 (21)	132 095 (77)	38 652 (23)
Sex
Female	502 055 (62)	408 138 (81)	93 917 (19)
Male	303 399 (38)	240 347 (79)	63 052 (21)
Race and ethnicity
African American or Black	115 030 (14)	89 989 (78)	25 041 (22)
Asian	32 715 (4)	28 914 (88)	3801 (12)
Hispanic	87 848 (11)	75 585 (86)	12 263 (14)
White	452 494 (56)	359 343 (79)	93 151 (21)
Unknown	117 367 (15)	94 654 (81)	22 713 (19)
Charlson Comorbidity Index score
0	402 074 (50)	347 405 (86)	54 669 (14)
1	181 334 (23)	142 311 (78)	39 023 (22)
≥2	222 046 (28)	158 769 (72)	63 277 (28)
Educational level
<12th grade	8876 (1)	7489 (84)	1387 (16)
High school	252 685 (31)	199 472 (79)	53 213 (21)
<Bachelor’s degree	400 954 (50)	323 922 (81)	77 032 (19)
Bachelor’s degree plus	131 882 (16)	108 625 (82)	23 257 (18)
Unknown	11 057 (1)	8977 (81)	2080 (19)
Census regionb
Midwest	100 631 (13)	81 156 (81)	19 475 (19)
Northeast	222 651 (28)	172 365 (77)	50 286 (23)
South	365 085 (45)	290 739 (80)	74 346 (20)
West	116 703 (15)	103 875 (89)	12 828 (11)
Site of initial presentation
Emergency department	185 338 (23)	119 600 (65)	65 738 (35)
Outpatient clinic	620 116 (77)	528 885 (85)	91 231 (15)

^a Percentage of row total population.

^b Less than 1% with unknown region.

Figure. Kaplan-Meier Neuroimaging Probability From Index Date to 6 Months by Clinical Setting of Diagnosis.

A head CT scan was used the most (ie, 47% of all tests [177 949 of 376 149] were head CT scans), followed by brain MRI (25% [93 130 of 376 149]), cerebrovascular USN (15% [56 175 of 376 149]), and magnetic resonance angiography (9% [34 026 of 376 149]) (Table 2). Multiple CT scans, multiple MRIs, or both CT scan and MRI were obtained for 12 289 of 156 969 (8%), 9197 of 156 969 (6%), and 2102 of 156 969 (5%) patients who underwent imaging, respectively.

Table 2. Neuroimaging Use and Direct Spending by Modality Among Commercial and Medicare Advantage Beneficiaries Within 6 Months of Initial Presentation for Dizziness.

Neuroimaging test	No. (%)	Total spending, $ (%)	Spending per test, median (range), $
			COM	MA
CT scan, head or brain, with or without contrast	177 949 (47)	14 569 788.39 (16)	68.97 (1.10-4600.00)	43.21 (1.17-2676.54)
CT scan, temporal bones, with or without contrast	2767 (1)	696 997.37 (1)	185.89 (3.42-1864.00)	167.85 (7.61-1065.75)
CT angiography	9666 (3)	1 643 720.48 (2)	143.51 (1.05-4055.33)	87.47 (1.94-2209.64)
MRI brain with or without contrast	93 130 (25)	52 072 344.74 (59)	319.63 (1.10-18 000.00)	122.99 (1.10-11 826.69)
MRI orbit, face, and neck with or without contrast	2436 (1)	1 460 882.80 (2)	303.75 (4.40-5494.92)	362.02 (43.04-2460.75)
MRA head and neck with or without contrast	34 026 (9)	8 197 024.41 (9)	123.09 (1.23-7535.00)	73.57 (4.11-6148.63)
Ultrasonography, cerebrovascular	56 175 (15)	10 005 288.84 (11)	132.42 (1.19-2046.00)	143.76 (1.74-3723.55)
Total	376 149	88 646 047.03	Mean (SD), 110.34 (413.12)	Mean (SD), 53.25 (199.36)

Abbreviations: COM, commercial insurance plans; CT, computed tomography; MA, Medicare Advantage plans; MRI, magnetic resonance imaging; MRA, magnetic resonance angiography.

Modality selection differed by initial presentation site and timing (Table 3). Within 6 months, the proportion of patients with dizziness who underwent a CT scan, MRI, or USN, respectively, was 32% (58 682 of 185 338), 4% (7743 of 185 338), and 5% (8786 of 185 338) after ED presentations and 6% (35 126 of 620 116), 8% (49 109 of 620 116), and 2% (13 287 of 620 116) after outpatient presentations. Among patients who received neuroimaging after ED presentation, a CT scan was the primary modality used both on the index date (92% [46 852 of 51 022]) and within 6 months (86% [56 682 of 65 738]). Among patients presenting to the outpatient clinic, relatively few (2% [12 698 of 620 116]) received imaging on the index date, most often CT (56% [7021 of 12 608]), and, over 6 months, a greater proportion of those who received imaging underwent MRI (54% [49 109 of 91 231]) compared with CT (39% [35 126 of 91 231]).

Table 3. Proportion of Individuals With Dizziness Undergoing Neuroimaging Modalities by Site of Initial Presentation and Time Point.

Modality	Individuals, No. (%)
	ED (n = 185 338)		Outpatient clinic (n = 620 116)		All sites (N = 805 454)
	Index date	Index date to 6 mo	Index date	Index date to 6 mo	Index date to 6 mo
CT scan	46 852 (25)	58 682 (32)	7021 (1)	35 126 (6)	93 808 (12)
MRI	6469 (4)	7743 (4)	2977 (1)	49 109 (8)	56 852 (7)
USN	2798 (2)	8786 (5)	3993 (1)	13 287 (2)	22 073 (3)
Anya	51 022 (28)	65 738 (36)	12 608 (2)	91 231 (15)	156 969 (20)

Abbreviations: CT, computed tomography, including computed tomography angiography; ED, emergency department; MRI, magnetic resonance imaging, including magnetic resonance angiography; USN, ultrasonography.

^a As individuals may receive more than 1 imaging test, the sum of the column totals is greater than the total number of individuals receiving any imaging modality.

Factors Associated With Neuroimaging Acquisition

In multivariable models (Table 4), any neuroimaging within 6 months of presentation for dizziness was significantly associated with age, higher Charlson Comorbidity Index (CCI) score, and initial ED rather than outpatient presentation. Asian and Hispanic patients were less likely than White patients to receive neuroimaging. In site-specific models (Table 4), advanced age had a stronger effect size on imaging acquisition after ED presentation compared with after outpatient presentations, whereas the CCI score had a stronger effect size in the outpatient setting than in the ED setting.

Table 4. Factors Associated With Neuroimaging Within 6 Months of Initial Presentation for Dizziness, by Site.

Factor	Odds Ratio (95% CI)
	All sites	ED	Outpatient
Age, y
18-39	1 [Reference]	1 [Reference]	1 [Reference]
40-49	1.33 (1.31-1.36)a	1.58 (1.53-1.63)a	1.22 (1.19-1.25)a
50-59	1.44 (1.42-1.47)a	1.86 (1.80-1.92)a	1.27 (1.24-1.30)a
60-64	1.47 (1.44-1.50)a	1.93 (1.86-2.01)a	1.27 (1.24-1.31)a
65-74	1.24 (1.20-1.27)a	1.53 (1.47-1.60)a	1.11 (1.07-1.14)a
≥75	1.51 (1.48-1.54)a	2.05 (1.98-2.11)a	1.30 (1.27-1.33)a
Gender
Male	1 [Reference]	1 [Reference]	1 [Reference]
Female	0.96 (0.94-0.97)a	0.99 (0.97-1.01)	0.94 (0.93-0.95)a
Race and ethnicity
White	1 [Reference]	1 [Reference]	1 [Reference]
African American or Black	0.97 (0.96-1.00)	0.96 (0.94-0.99)a	1.00 (0.98-1.02)
Asian	0.58 (0.56-0.60)a	0.64 (0.60-0.68)a	0.56 (0.53-0.58)a
Hispanic	0.66 (0.64-0.67)a	0.65 (0.63-0.68)a	0.66 (0.64-0.67)a
Unknown	0.96 (0.94-0.97)a	0.95 (0.93-0.98)a	0.96 (0.94-0.98)a
Charlson Comorbidity Index Score
0	1 [Reference]	1 [Reference]	1 [Reference]
1	1.59 (1.57-1.62)a	1.30 (1.27-1.34)a	1.81 (1.77-1.84)a
≥2	2.08 (2.04-2.11)a	1.45 (1.41-1.48)a	2.61 (2.57-2.66)a
Site
Outpatient	1 [Reference]	NA	NA
ED	3.07 (3.03-3.11)a	NA	NA

Abbreviations: ED, emergency department; NA, not applicable.

^a Statistically significant.

In a Cox proportional hazards regression model restricted to the outpatient setting, the risk of neuroimaging was higher for patients first seeing a neurology clinician (hazard ratio [HR], 1.60 [95% CI, 1.56-1.64]) or a cardiology clinician (HR, 1.03 [95% CI, 1.04-1.06]) compared with those first seeing a primary care clinician, and the risk of neuroimaging was lower for patients first seeing an otolaryngology clinician (HR, 0.92 [95% CI, 0.89-0.94]) compared with those first seeing a primary care clinician. Older age (≥75 years vs 18-39 years: HR, 1.06 [95% CI, 1.03-1.08]) and CCI score (CCI score 1 vs CCI score 0: HR, 1.11 [95% CI, 1.09-1.13]; CCI score ≥2 vs CCI score 0: HR, 1.04 [95% CI, 1.02-1.06]) had a modestly higher risk of a neuroimaging event, whereas women had a lower risk of a neuroimaging event than men (HR, 0.91 [95% CI, 0.90-0.92], vs men), and Asian patients (HR, 0.92 [95% CI, 0.89-0.92]) and Hispanic patients (HR, 0.92 [95% CI, 0.89-0.94]) had a lower risk of a neuroimaging event than White patients.

Spending

Neuroimaging spending is detailed in Table 2. Of $88 646 047.03 in total neuroimaging spending (patient spending plus amounts paid by insurance plan), MRI or magnetic resonance angiography accounted for 70% ($61 730 251.95), CT scan or CT angiography accounted for 19% ($16 910 506.24), and USN studies accounted for 11% ($10 005 288.84). Although spending ranges were right skewed, per-test median total spending ranged from $68.97 (CT scan of the head) to $319.63 (MRI of the brain brain) among commercially insured individuals and from $43.21 (CT scan of the head) to $362.02 (MRI of the orbit, face, and neck) among individuals with MA plans.

Discussion

This study used national administrative claims data from commercial insurance and managed care MA plans to quantify the use of neuroimaging across ambulatory and ED settings. Among 805 454 patients presenting with dizziness, 20% underwent neuroimaging within 6 months, 26% of whom underwent 2 or more neuroimaging tests. When neuroimaging was acquired urgently, a head CT scan was the modality most used across settings. However, within 6 months, MRI was performed in a greater proportion of outpatient clinic than ED presentations and accounted for the most neuroimaging spending overall, albeit at a lower-than-expected per-test cost. This claims-based analysis was designed to address the limitations of previous examinations of neuroimaging for patients with dizziness that were restricted to single time points, care settings, and clinician specialties or subject to patient recall bias and that did not capture actual spending.^1,4,23,24,25 By considering care for dizziness across heterogeneous settings and times, our analyses revealed patterns of neuroimaging use and spending that could inform clinical and policy interventions intended to promote high-value diagnostic practices.

There has been an emphasis placed on optimizing neuroimaging use for acute dizziness in ED settings, where urgency is associated with reduced in-the-moment payer and health system scrutiny. The ED neuroimaging rates that we observed (28% at presentation and 36% over 6 months) are consistent with the prior estimates of 18% to 48% from US health system ED records and the National Hospital Ambulatory Medical Care Survey.^{1,24,25,26,27,28} These studies and our study documented reliance on the insensitive head CT scan in the urgent time frame, which is concerning because 35% of strokes among patients presenting to the ED with dizziness are missed.²⁹ Among ED presentations for dizziness, only 0.6% to 3.6% of CT scans and 3.9% to 12.2% of MRI scans revealed clinically significant findings.^{24,25,27,30,31,32} Morbidity from missed pathology, particularly stroke, may justify imaging regardless of yield,³³ but higher use of neuroimaging was not associated with higher yields.^34,35

Emergency department clinician behavior surrounding imaging case selection is a rational focus of efforts to improve yield and reduce missed pathology. This focus is undergirded by policy, as the Emergency Medical Treatment and Active Labor Act and the Patient Protection and Affordable Care Act ensure that patients with severe symptoms seldom are denied acute care and patients receive ED care as determined by their evaluating practitioners rather than insurance status or ability to pay. In particular, ED clinicians are encouraged to attend to the case presentation of a patient with dizziness, because the central causes are most prevalent among individuals with neurologic symptoms and signs, including imbalance or gait instability, sudden hearing loss, headache, visual disturbance, paresthesia, and paralysis.^25,27,36 For patients with acute vestibular syndrome, characterized by continuous severe vertigo or dizziness with nystagmus and gait unsteadiness, the Head Impulse, Nystagmus, Test of Skew (HINTS) examination helps differentiate stroke from its “benign” counterpart, vestibular neuritis.^37,38 Normal horizontal head impulse, direction-changing nystagmus, or skew deviation suggest central causes and need for imaging.^39,40 The HINTS examination had a pooled sensitivity of 96% and a pooled specificity of 71% to detect stroke, surpassing early MRI.⁴¹ In our study, ED imaging decisions appeared to be associated with perceived cerebrovascular risk (age and comorbidities)^24,27 and race and ethnicity. Neuroimaging findings are associated with age,⁴² but multiple studies suggest that comorbidities alone are insufficient to drive case selection.^{25,27,36,42,43,44} To meet performance standards for stroke center certification by The Joint Commission, EDs may implement processes that direct patients with stroke symptoms to undergo a CT scan on arrival before full examination to minimize door-to-needle time for thrombolytic administration.^45,46 In these instances, CT scan acquisition for dizziness may be driven as strongly by quality improvement protocols as by medical decision-making. The patient characteristics associated with the lowest risk of imaging in our study (younger age, fewer comorbidities, and Hispanic or Asian race and ethnicity) correspond to those portending a higher risk of missed stroke at ED presentation for dizziness,⁴⁷ suggesting that there is more targeted work to be done. As access to an around-the-clock neurotologic expert examination is limited, new point-of-care testing paradigms that incorporate video-oculography for rapid patient triage are under investigation and are promising means of curbing inappropriate neuroimaging.⁴⁸

Reconsideration of the use of neuroimaging at outpatient presentation is also warranted based on the patterns observed among the 77% of individuals with dizziness presenting to outpatient clinics. The overall imaging rate (15%) was lower after outpatient presentations than after ED presentations, but 1.4 times as many actual patients underwent imaging after outpatient presentations. In a National Ambulatory Medical Care Survey analysis, 5.5% of adult outpatient visits for dizziness resulted in imaging orders.⁴ Although impending strokes may be missed,⁴⁹ several studies indicate that neuroimaging has a lower yield and therapeutic efficacy in ambulatory clinics where central lesion prevalence is lower and presentations are more often episodic or chronic than acute.⁵⁰ For example, less than 3% of brain MRI scans or CT angiography scans yielded new findings among outpatients with isolated dizziness, just over half of which altered management.⁵¹ When patients also had abnormal vestibular testing results, 5.5% of MRI scans yielded treatable causative lesions, and all had additional symptoms.⁵² In our cohort, MRI was more often used after outpatient presentations than after ED presentations, potentially evidencing greater knowledge dissemination or implementation regarding MRI’s diagnostic superiority and/or technology access. However, 39% of outpatients who received imaging recevied a CT scan, and, because temporal bone CT scans comprised only 2% (2767 of 180 716) of the CT scans performed, concern for central lesions rather than labyrinthine pathology was likely the reason for most decisions. Similarly, although carotid stenosis is an uncommon dizziness precipitant and dizziness has been highlighted as an inappropriate indication for carotid USN that precipitates unnecessary revascularization,⁵³ 15% of those who received imaging after outpatient presentation underwent USN.

This study revealed a window of opportunity for the use of productive interventions surrounding outpatient case selection. Decisions to order imaging tests decreased relatively early in the outpatient diagnostic pathway, with a median of 10 days to testing. Outpatient clinic clinicians may feel pressured to order imaging tests defensively to rule out diagnoses rather than wait for the complex evolution of symptoms prior to initiating treatment. Although outpatient clinicians may also less frequently encounter opportunities to use the HINTS examination, targeted clinical education may be warranted because missed stroke-related harms after primary care presentations for dizziness are twice that of specialty care presentations.⁴⁹ For episodic and chronic presentations, imaging case selection and appropriate use could be informed by novel decision support tools built into the electronic ordering systems based on contemporary diagnostic criteria^54,55,56 and practice guidelines.^57,58 Clinicians may also benefit from education on the use of asynchronous virtual examinations of patients via cell phone video recordings of paroxysmal attacks because the findings (eg, central vs peripheral) may inform decisions on the timing, setting, and need for neuroimaging during an in-person evaluation.⁵⁹ Machine learning is being used to develop and implement algorithms with negative predictive value for central lesions⁶⁰ and to discriminate between peripheral diseases.⁶¹ Refined algorithms may decrease the use of unnecessary neuroimaging in a manner less dependent on clinician education. Additional possible levers of influence on test-ordering behavior for patients with dizziness could include the incorporation of imaging into incentivized quality metrics and required clinical documentation for prior authorization.

Cost is a key factor in assessing the value of neuroimaging for dizziness, but accurate estimates have been lacking. Unlike prior analyses, our data set afforded documentation of neuroimaging spending (patient out-of-pocket costs plus amounts paid by commerical and MA plans). We observed median spending of less than $69, $320, and $144 per CT scan, MRI, and USN, respectively. Prior estimates (circa 2011) were substantially higher because they imputed private sector charge tables (CT scan, $1220-1856; MRI, $2696-$4029^24,25) or fee-for-service Medicare fee schedules (CT scan, $236; MRI, $1220¹). Lower-than-expected per-test spending may also occur in other plans over time, particularly as fee-for-service Medicare reimbursement decreased from 2007 to 2019 for all imaging modalities (for MRI most of all).⁶² Prior ED-restricted analyses estimated higher CT scan costs than MRI costs,¹ which were associated with CT overuse.^1,24,25 Even with lower-than-anticipated spending per test, our inclusion of outpatient presentations and fuller episodes of care demonstrated that total MRI spending was 3.5 times CT scan spending. This finding further emphasizes that efforts to rightsize imaging use for dizziness should strongly consider appropriate MRI use during outpatient and post-ED care.

Given recent policy initiatives, lower-than-expected per-test neuroimaging costs may influence patterns of care. For several decades, health policy makers have argued for greater transparency of prices and quality. The US Department of Health and Human Services mandated hospitals to post list prices, or chargemasters, for services as of January 1, 2019. Owing to legislation that began in early 2021, hospitals are required to post payer-specific negotiated charges and discounted cash prices in machine-readable files and to disclose these charges for 300 “shoppable” services, 70 of which were specified by the Center for Medicare & Medicaid Services (CMS).^16,17 Thirteen of the 70 CMS-specified shoppable services are imaging procedures, including CT and MRI. It is unclear how patients, clinicians, and health plans will respond to price transparency.⁶³ Legislated price transparency assumes that patients will be encouraged to pursue the lowest price and that the resultant market-based price competition will force lower prices.⁶⁴ With seemingly low fees, whether from market competition or plans, those who desire neuroimaging may be even more likely to pursue it. Thus, while health plans aim to lower costs, informed and potentially anxious patients may stimulate the overuse of neuroimaging for dizziness, especially in outpatient settings that provide more time to shop.⁶³ To limit the unwarranted adoption of patient-driven imaging tests, there may be a need for dizziness care to be a focus of campaigns, such as Choosing Wisely, that engage patients in clinical discussions regarding testing and treatments with the goal of encouraging care that is supported by evidence and is truly necessary.⁶⁵

Limitations

This study has several limitations. The population represents US adults with commercial insurance or MA plans. Commercial beneficiaries in the OptumLabs Data Warehouse have been shown to be similar in age, race and ethnicity, and sex to the broader US commercially insured population,⁶⁶ but the findings may not be generalizable to plans not included in the OptumLabs Data Warehouse,⁶⁷ fee-for-service Medicare, Medicaid, or uninsured individuals. Because clinicians may not assign dizziness codes for all symptomatic individuals, cohort selection had greater specificity than sensitivity across settings. Emergency department imaging estimates were consistent with published work, lending credence to our broader findings. Claims were anchored to presentation date to capture neuroimaging associated with dizziness during the episode of care, but some tests may have had other indications or may have predated the initial dizziness consultation. We did not determine whether patients who presented to outpatient settings later went to EDs or vice versa. Clinician characteristics, such as National Provider Identifier or number of years in practice, were unavailable to observe granularity beyond specialty. Despite these limitations, we believe our results emphasize important patterns of neuroimaging use in ambulatory and acute care settings.

Conclusions

The findings of this cross-sectional study suggest that both clinicians and patients can be prone to use noninvasive tests for dizziness, and, without clearly defined clinical pathways and with technologies widely available, imaging may be an appealing high-end modality. However, routine use of neuroimaging, even of the more sensitive MRI, for dizziness is unlikely to be helpful in many cases, and the rates of imaging that we observed appear higher than necessary. Through clinician-, payer-, and policy-level interventions, avoidance of unnecessary imaging tests that are low yield and carry a risk could lower resource use and cost. There appears to be a particular opportunity to improve value by addressing the allowable indications for MRI use outside the ED. Future work should focus on creating a guideline-concordant, policy-supported clinical toolkit with interventions targeted at the time of initial presentation to improve care.

Supplement.

eFigure. CONSORT Diagram

eTable. Current Procedural Technology (CPT) Codes for Neuroimaging Studies