Usefulness of exam questions and vital signs for predicting the outcome of objective vestibular tests

Abstract

Objective:

to determine the value of standard clinic screening questions and vital signs in predicting abnormal vestibular function, indicated by standard objective diagnostic tests.

Methods:

We reviewed electronic medical records of 150 patients seen by the neurotologists or the physician assistant they supervised, in an out-patient tertiary care clinic, between June 2018 and March 2020, and subsequently referred for the complete objective vestibular test battery (VB).

Results:

Of standard questions asked during the initial exam about vertigo, disequilibrium, lightheadedness and oscillopsia, only vertigo predicted an abnormal response on the VB. More males than females had abnormal VB responses, p<0.05. Pulse was not related to VB score. Significantly more subjects with blood pressure in the range for Stage 2 hypertension (BP Stage 2) had abnormal than normal results on the VB, p<0.00001. Subjects with BP Stage 2 had high rates of diabetes (34.2%) and hypertension (68.4%) as diagnosed by their primary care physicians or cardiologists.

Conclusion:

Complaints of subjective vertigo and BP in the range of hypertension stage 2 are most likely to predict abnormal findings on the VB. Therefore, during an examination of a patient who comes in complaining of dizziness, two measures may be the most useful for screening: BP in the range of hypertension type 2, when BP is taken by a nurse, and a question to determine whether or not the patient has true vertigo.

Introduction

Patients with vestibular disorders often complain of vague symptoms. Sometimes, the condition that might present as a possible vestibular impairment is really caused by a different issue. Therefore, during the initial neurotology examination providers take several measures intended to dissect the nature of a patient’s symptom profile. The true, independent value of some of these measures is not clear. For example, at the beginning of any outpatient clinic visit, standard practice includes taking vital signs, but those data are not always incorporated into the diagnostic algorithm for dizziness.

Patients who complain of dizziness can be challenging for even the most skilled clinician. The physician or physician assistant (MD/PA) typically asks each patient questions about the history and current complaint or reason for the visit. At our institution, when the patient has informed the schedulers that the reason for the visit is dizziness or imbalance, the MD/PA typically asks questions about signs and symptoms of vestibular disorders. For example, experts recommend asking questions about the characteristics of the bothersome sensation, such as vertigo or lightheadedness, the time course, the associated signs and symptoms and exacerbating factors.¹

Care for dizzy patients might be improved at no additional expense if we knew more about the value of specific information routinely gathered during the initial visit. Therefore, the goal of the present study was to determine if some vital signs or patient-reported symptoms predict subsequent performance on objective diagnostic tests of the vestibular system. We tested the hypothesis that elevated blood pressure and female sex were the best predictors.

Methods

The data for this study were collected from existing electronic medical records for patients seen between July 2018 and March 2020. Records from 150 consecutive patients with complete records were used. The sample included 63 males and 87 females, aged 8.6 to 89.7, mean 55.5 (SD 15.5). See Table 1 for demographic details. This study was approved by the Institutional Review Board for Human Subjects Research for Baylor College of Medicine and Affiliated Hospitals.

Table 1.

Demographics and relationship to responses on the VB.

Patient characteristics	Abnormal VB (N=81)	Normal VB (N=69)	p-value
Age (yrs), (SD)	60.8 (12.7)	49.5 (16.3)	< 0.0001
Mean (SD) number of days between clinic visit and VB	35.8 (42.7)	37.8 (66.8)	0.47
Sex: Female, Number (% of sample)	41 (50.6%)	46 (66.7%)	0.047
Male, Number (% of sample)	40 (49.8%)	23 (33.3%)

Measures

At the beginning of each visit, after the patient sat quietly in the examination room for at least 5 minutes, a medical assistant took the patient’s blood pressure (BP) and pulse with a digital sphygmomanometer (WelchAllyn, Nellcor Sp0₂ Spot Vital Signs device). If a patient had abnormalities in these vital signs, a nurse retested the vital signs after the patient sat quietly for at least 5 more minutes. Due to the security configuration in our building all patients took the elevator to our floor. No visitors entering the building can take the stairs. When they arrived on our floor, after checking in at the reception desk patients typically sat and waited before being escorted to the examination room by a medical assistant. Thus, patients were not subjected to significant physical exertion prior to vital signs being taken.

The medical assistant updated the patient’s medical history using a standard form in the electronic medical record. Therefore, we were able to determine if subjects reported having a history of hypertension or diabetes. If subjects are seen by primary care or cardiology at our institution the medical assistant was able to update the medical record based on those findings.

After the vital signs were taken and the medical history was updated, the patient was seen by a neurotologist or the physician assistant on the neurotology service, who all asked questions about vertigo, oscillopsia, disequilibrium, and lightheadedness. They also asked other questions that were germane to each individual patient and they performed clinical examinations. Those other questions and results of the clinical examination were not included in this study because of the variation in other questions and clinical exam procedures.

The standard clinical objective vestibular test battery (VB) at this institution has several components. 1) Cervical vestibular evoked myogenic potentials are recorded with surface Ag/AgCl electrodes paced over the bulk of the sternocleidomastoid halfway between the mastoid tip and the sternal notch (reference electrode is placed over the sternum and the ground electrode is placed on the forehead) while the patient lies supine and lifts his head against gravity. Tone bursts are presented through earphones at a central frequency of 500 Hz, maximum 100 dB nHL, 200 msec interval testing. Initial stimuli are provided at 90 dBHL, decreasing in 5 dBHL steps. The threshold is determined as the lowest stimulus level still producing a response.

2) In the rotatory chair (Neurokinetics) patients’ eye movements are recorded with infrared video oculography (VOG). They have several tests: a) spontaneous nystagmus in darkness for 10 seconds. Three or more beats of nystagmus indicate an abnormal response. b) Low frequency sinusoidal tests of the vestibulo-ocular reflex at 0.01 Hz, 0.08 Hz, 0.32 Hz and 0.64 Hz, ± 60°/sec, using norms for gain and phase provided by the manufacturer.

3) Dix-Hallpike maneuvers and supine roll tests are recorded with VOG. Three or more beats of nystagmus indicate an abnormal response. 4) Bi-thermal caloric tests with water for 40 sec per irrigation, at 30° C and 43°C, with the neck flexed 30°, also recorded with VOG. An abnormal caloric test is defined as ≥ 25% unilateral weakness. Other oculomotor tests, not included in this analysis, are also included in the VB.

For the purpose of this study, any abnormal response on vestibular testing was coded in the study database as positive. Individual diagnoses, if available, were not included due to lack of clarity in diagnostic codes.

Statistical methods

If vital signs were taken twice, the second time by the nurse, then the statistical analyses used the values recorded by the nurse. Patients with an abnormal VB result were compared to those with normal VB results by Chi-square/Fisher exact tests for categorical data, t-tests for normally distributed data and Wilcoxon rank sum tests for non-normal data. The relationship between patients’ responses to physicians’ and technicians’ questions were assessed by Kappa statistics. P<0.05 was considered significant. All analyses were performed in SAS statistical software (version 9.4, Cary, NC).

Results

Patients with abnormal VB results were significantly older than those with normal results, p< 0.0001. Sex was also related to test outcomes, with males more likely than females to have abnormal responses, p=0.047. The number of days between the initial neurotology exam and the VB was not related. See Table 1.

Of the four MD/PA questions, only responses to the question about vertigo predicted an abnormal response on the VB, p=0.015. Responses to questions about oscillopsia, disequilibrium and lightheadedness were not predictive. See Table 2.

Table 2.

The number (and percent) of Yes responses to questions asked by the physician or physician assistant (MD) and by the technician (tech), by abnormal and normal responses on the VB.

Question, % Yes	Abnormal VB	Normal VB	p-value
MD: Vertigo	64 (79%)	42 (60.9%)	0.015
MD: Oscillopsia	27 (33.3%)	14 (20.3%)	0.07
MD: Disequilibrium	66 (81.5%)	53 (76.8%)	0.48
MD: Lightheadedness	36 (44.4%)	31 (44.9%)	0.95
Tech: Vertigo	39 (59.0%)	40 (57.9%)	0.33
Tech: Oscillopsia	47 (59.5%)	38 (55.1%)	0.59
Tech: Disequilibrium	80 (98.8%)	67 (97.1%)	0.59
Tech: Lightheadedness	55 (68.8%)	56 (81.2%)	0.08

Many patients visit our clinics complaining of vague symptoms of “dizziness. ” Some of those complaints may be cardiovascular in nature. Therefore, we tested the relationships between vital signs, i.e. BP and pulse rate, and the responses to MD/PA questions, the response to the technician question about positional changes, and the outcome of the VB. Vital signs were coded in the database using the published ranges for normal BP, hypotension, elevated BP, Stage 1 hypertension, Stage 2 hypertension, and abnormal pulse rate.² We used the following definitions: normal BP, < 120/80 mm HG, hypotension < 90/ < 60 mm HG, Stage 1 hypertension, 130 to 139/80 to 89 mm HG, Stage 2 hypertension, ≥ 140/≥ 90 mm Hg. (No patients had BP in the intermediate range of elevated BP, 120 to 128 mm HG/ < 80 mm HG.) We used the range of 60 to 100 beats per minute as the normal range of pulse.

Overall, frequency of abnormal BP was greater among patients with abnormal than normal VB results (85% vs 76%), specifically for the Stage 2 hypertension subtype (47.5% vs 14.9% p<.0001). See Table 3. No interaction was found between patient’s age and BP in the range of stage 2 hypertension (p for interaction=0.69). Also, when systolic and diastolic BP were examined separately, both measures were significantly higher in subjects with abnormal than normal VB responses. We found no differences in abnormal or normal VB group by pulse rate. See Table 3.

able 3.

Number (and percentage) of subjects who had one or more abnormal and normal VB scores, by type of blood pressure, and by type of pulse rate.

Test	VB Result		p-value
	Abnormal VB	Normal VB
Blood pressure
Normal BP	12 (15%)	16 (24%)	(reference)
Abnormal BP, any type	68 (85%)	51 (76%)	0.18
Hypotension	1 (1.3%)	3 (4.5%)	0.63
Elevated BP	10 (12.5%)	12 (17.9%)	0.85
Hypertension Stage 1	19 (23.8%)	26 (38.8%)	0.95
Hypertension Stage 2	38 (47.5%)	10 (14.9%)	0.001
Systolic BP, Mean (SD)	133.5 (16.9)	126.7 (14.5)	0.01
Diastolic BP, Mean (SD)	81.7 (9.8)	78.4 (10.5)	0.04
Pulse rate
Normal	69 (86.2%)	60 (89.6%)	0.45
Decreased	9 (11.3%)	4 (5.9%)
Increased	2 (2.5%)	3(4.5%)

Males and females did not differ significantly in the frequency of blood pressure in the ranges for Hypertension Stages 1 and 2, and they did not differ by stage. We found 24 females (28.6%) and 21 males (33.3%) with BP in the range of Hypertension Stage 1. We found 26 females (31.0%) and 22 males (34.9%) with BP in the range of Hypertension Stage 2.

BP in the range of hypertension stages 1 or 2 could have been related to diagnoses of hypertension or diabetes. To determine which patients had diagnosed hypertension or diabetes we used subject-supplied health history and information from the electronic medical record if the subject was followed at our institution for primary care or cardiology. Of subjects with abnormal VB results and BP in the range of stage 2 hypertension, 26 (68.4%) had been diagnosed with hypertension and 13 (34.2%) had been diagnosed with diabetes.

Finally, we examined the utility of combined variables that were individually associated with VB. A model that included patient’s age (≥60 vs <60), gender, vertigo and oscillopsia questions had an AUC of 0.75 for predicting abnormal VB result (acceptable discrimination). Using the same model but with the addition of blood pressure in the range of stage 2 hypertension had an AUC of 0.80 (excellent) attesting to the usefulness of blood pressure in the range stage 2 hypertension for predicting abnormal VB.

Discussion

The main finding from this study is that some measures taken during the initial neurotology visit may be useful for screening patients for vestibular disorders. Other measures are not clearly useful for that purpose. These data indicate that having elevated BP in the range of Stage 2 hypertension is related to abnormal VB results. This finding is consistent with the idea that some patients who are seen in the neurotology clinic for dizziness may have underlying microvascular lesions that could cause vestibular impairments. Thus, BP in the range of Stage 2 hypertension may be a risk factor for vestibular disorders.

Patients were not subjected to undue physical stress immediately prior to the visit. We could not determine traffic conditions, levels of personal stress, stress due to “white coat syndrome” -- i.e. merely being in a clinical environment, or whether or not patients had taken antihypertensive medication. These factors, though, were likely to have been equally distributed across patient subgroups. Clinical protocols for BP measurement, allowing for patients to sit quietly in between measurements, probably controlled for some confounding factors.

The MD/PA question about true vertigo is useful for predicting response to the VB. The other MD/PA questions, however, are not useful for the purpose of screening for the need for the VB. The questions about “dizziness” and balance may be not useful because accurately describing these problems is so complicated. For example, while many patients may initially describe a balance problem as “dizziness” upon further discussion during the visit they may clarify by describing syncope, unsteadiness due to pain while ambulating, or anxiety. This finding supports previous evidence that patients complaining of “dizziness” of various types tend to confuse them when responding to questions.³ Thus, clinicians could be more efficient in the interview with the patient, without compromising patient care, by limiting certain types of questions.

The finding that being male was related to having an abnormal VB result was surprising. This finding may represent sociological phenomena that influence the patient’s decision to seek care. For example Sweeny et al. have shown that a spouse may influence factors related to vestibular schwannoma diagnosis and management.⁴ Also, some evidence suggests that the desire to improve participation in personally meaningful occupations -- such as self-care, professional activities, family responsibilities, or preferred sports --is a primary reason for patients to seek care and influences their response to therapy.⁵ Extrapolating those findings to the present study, perhaps some older male patients wait to seek care until either the ability to participate in personally meaningful activities is compromised or the patient’s or partner encourages him to visit the otolaryngology clinic. A clinician who sees dizzy patients should consider these issues during treatment planning.

The descriptive data show high percentages of subjects with known hypertension and diabetes who have abnormal vestibular test findings. This finding may be consistent with microvascular-related injuries in the vestibular system. Perhaps, if hypertension is not well controlled, VB test results may be affected. Whether or not a clinician my want to delay VB or other testing until stage 2 hypertension is better controlled is not clear. Out data suggest that providers who may not focus on hypertension management should, nevertheless, reinforce the importance of BP control, even if only through encouragement to see the primary care physician or cardiologist when patients appear to have hypertension. Also, the likelihood of having vestibular problems associated with uncontrolled hypertension is a good reason for the otolaryngologist to be in touch with other members of the patient’s medical team and for the otolaryngologist to help identify the probable cause of the patient’s initial complaint of “dizziness”. The ideal means for communicating results and recommendations of this sort is unclear, and beyond the scope of this study. The otolaryngology nurse may be designated to reach out to the primary care physician or cardiologist when uncontrolled hypertension is suspected.

In the current health care environment, clinic visits are necessarily short. Many visits are done via telemedicine. For the busy clinician, the most predictive and most useful of the screening measures that we examined is the question about vertigo – if the clinician clearly defines vertigo – and very high BP. Many patients these days take their own BP with home-based equipment. For screening, clinicians may find that these measures help predict the outcome of the VB, without resorting to a long list of other questions and clinic measures.

This study had several limitations. It was retrospective. Therefore, we were not able to interview subjects about their medications and reactions to medications. We did not have complete medical records on all subjects. Therefore, information that primary care physicians and cardiologists might have collected about medications was not available to us. Similarly, we lacked definitive information with which to rule out or make diagnoses of Stage 2 hypertension for many patients. Although these issues limited the extent of the study, they also provide a realistic model of the situation faced by many clinicians.

Conclusion

During the initial visit with a dizzy patient otolaryngologists can be more efficient by limiting some of their interview questions. Standard vital sign measurement may reveal elevated blood pressure in the range of Stage 2 hypertension. Such high blood pressure may predict impaired vestibular function, thus providing the physician with another tool to aide in clinical decision making about the need for extensive objective vestibular testing. Such high blood pressure may also indicate the presence of other health problems and might indicate a reason for outreach to other members of the medical team. Awareness of these issues may lead to improved patient care.