Canalith repositioning procedure illustrated for involvement of the right posterior semicircular canal. For positions B through D, if nystagmus was observed, the provoking position was maintained until the nystagmus stopped. If no nystagmus was observed, the position was maintained for 30 seconds.
The rate of recurrence of benign paroxysmal positional vertigo symptoms was determined on the basis of follow-up of 70 patients.
Kaplan-Meier estimation of time to recurrence for the vibration and no-vibration groups.
Hain TC, Helminski JO, Reis IL, Uddin MK. Vibration Does Not Improve Results of the Canalith Repositioning Procedure. Arch Otolaryngol Head Neck Surg. 2000;126(5):617-622. doi:10.1001/archotol.126.5.617
To determine whether, in patients with benign paroxysmal positional vertigo (BPPV), the canalith repositioning procedure performed with vibration applied over the mastoid bone of the affected ear is more effective in resolving the symptoms and preventing recurrence of BPPV than the procedure performed without vibration.
Retrospective case review.
Tertiary referral center.
Ninety-four patients diagnosed as having BPPV involving the posterior semicircular canal.
Patients were assigned to one of 2 treatment groups: the canalith repositioning procedure with vibration (n=44) and with no vibration (n=50).
Main Outcome Measures
Effectiveness of treatment was determined through clinical reevaluation or reported through a telephone interview 1 week after treatment. Intensity of symptoms was quantified on a scale of 1 to 3 (mild, moderate, or severe); effectiveness of treatment was categorized on a scale of 1 to 4 (cure, much better, better, or no change). Rate of recurrence was determined through later clinical reevaluation or a telephone interview.
At 1 week, 57 of the 94 patients were cured and 16 were much better, providing a 78% overall success rate. There was no significant difference in effectiveness of the treatment or the frequency of reoccurrence of BPPV between the vibration and no-vibration groups as determined from the Kaplan-Meier product-limit method and log-rank test. Rate of recurrence was 47% at a maximum follow-up of 5.25 years.
Our results suggest that, while the canalith repositioning procedure is effective in the treatment of BPPV, vibration applied during the maneuver does not significantly affect short-term or long-term outcomes.
BENIGN PAROXYSMAL positional vertigo (BPPV) is the most common cause of vertigo in the elderly.1,2 It is characterized by a brief period of vertigo experienced when the position of the patient's head is changed relative to gravity. The vertigo is caused by abnormal mechanical stimulation of the dependent posterior semicircular canal. Recently, there have been dramatic improvements in the treatment of BPPV. Before 1980, patients were treated with medications and advised to avoid moving their heads into positions that provoked vertigo. However, with this management, only 25% of patients with BPPV were symptom free within 3 months.3 A better understanding of the cause of BPPV4 led to the development of positional exercises. In 1980, Brandt and Daroff5 introduced the first positional exercises for the treatment of BPPV. Patients performed Brandt-Daroff exercises throughout the day until no symptoms of vertigo were experienced. Ninety-eight percent of patients were cured within 1 to 2 weeks. The Brandt-Daroff exercises were often impractical because patients did not tolerate repeated provocation of symptoms. Since then, other positional maneuvers have been introduced that may eliminate BPPV within a single treatment session.6- 8 The canalith repositioning procedure6 is the most commonly used positional maneuver in the United States.
The canalith repositioning procedure is illustrated in Figure 1. The procedure is based on a hypothetical mechanism, "canalithiasis,"6 which denotes free debris or displaced otoconia within the long arm of the posterior semicircular canal.4 When the position of the patient's head is changed relative to gravity, movement of the debris causes abnormal stimulation of the semicircular canal. Such debris has been visualized at the time of surgery in patients with BPPV.9 The aim of the canalith repositioning procedure is to move the debris from the posterior semicircular canal into the vestibule.6
Epley6 suggested that vibration should be applied over the mastoid of the affected ear during the canalith repositioning procedure to loosen debris that marginated within the semicircular canals. The maneuver is repeated as many as 6 times within a treatment session until no nystagmus is observed during the last cycle or until no progress is evident in the last 2 cycles. The maneuver is repeated weekly until vertigo has resolved and the Hallpike maneuver is negative.6 Fifty-seven percent to 100% of patients treated are cured of BPPV.6,8,10- 13 Variation in the effectiveness of treatment may be due to modifications made to the procedure, differences in the definition of cure, and differences in follow-up.
It is important to determine whether modifications made to the canalith repositioning procedure influence the outcome. One common modification is to perform the procedure without the use of vibration.11,13- 17 Recently, Li16 compared the efficacy of treatment between the canalith repositioning procedure performed with and without vibration and reported that 70% of the vibration group were cured, while none of the no-vibration group was cured. He concluded that vibration was "critical to the success" of the procedure. However, reliable inferences cannot be made from the Li study because of the small number of subjects (27 with vibration, 10 without). In addition, the results of Li's no-vibration group differ from the results of others who performed the procedure without vibration (Table 1). Again, this suggests that Li's sample size was inadequate. For these reasons, we examined a larger patient population to determine whether vibration applied over the mastoid bone of the affected ear improves the outcome of the canalith repositioning procedure.
As part of a neuro-otologic examination, the diagnosis of BPPV was established on the basis of the patient's response to the Dix-Hallpike maneuver. Three criteria were required for a diagnosis of BPPV: (1) a 1- to 20-second latency before the onset of vertigo and nystagmus, (2) observation of a rotatory and/or upward-directed nystagmus, and (3) vertigo and nystagmus of less than 60 seconds in duration. Patients were excluded from the study if the diagnosis of bilateral BPPV was established or if central nervous system involvement was identified on the basis of history, magnetic resonance imaging, or neurological examination.
Patients were treated with the canalith repositioning procedure without vibration from 1991 to 1997 and with vibration from 1994 to 1996. The canalith repositioning procedure, illustrated in Figure 1 for involvement of the right posterior semicircular canal, consisted of the clinician moving the patient through a series of 5 positions (Figure 1, A-E). The procedure was performed with the patient on an examination table. The table was adjusted so that the head was tilted posteriorly 20° from the trunk, when the patient was lying supine.
To begin, the patient was seated and the head was rotated 45° toward the right (Figure 1, A). The patient was rapidly lowered into the right head-hanging position of the Dix-Hallpike maneuver (Figure 1, B). If nystagmus was observed, the provoking position was maintained until the nystagmus stopped; if no nystagmus was observed, the position was maintained for 30 seconds. In the vibration group, once the position was assumed, vibration was applied with a handheld oscillator (Pollonex Aquassager, model K120; Holmes Corp, Sedalia, Mo) with a frequency of approximately 85 Hz over the mastoid area of the involved side for 20 seconds. This was the same device used by Li.16 If a burst of nystagmus was observed, the position was maintained until the nystagmus ceased. In the no-vibration group, vibration was not applied.
While the head was tilted back, the head was quickly turned 45° toward the left into the left head-hanging position of the Dix-Hallpike maneuver (Figure 1, C). The timing and application of the vibration were the same as described in Figure 1, B.
The patient was rolled into the left side-lying position (Figure 1, D), keeping the head-on-neck position as shown in Figure 1, C, to obtain a 180° position of the head with respect to Figure 1, B. The timing and application of the vibration were the same as described in Figure 1, B.
The patient was returned rapidly to the sitting position, keeping the same head-on-neck position as in Figure 1, C, until entirely upright (Figure 1, E). Once the patient was upright, the head was flexed slightly forward. The position was maintained for 1 minute. The positions in Figure 1 were repeated until the procedure had been performed a total of 3 times. For both groups, verbal and written instructions were given to the patient to sleep semirecumbent during the next 48 hours. For the remainder of the week, they were asked to avoid the following: sleeping with the ear in a dependent position, rapid head movements, extreme flexion and extension of the neck, and provoking positions.
One week after treatment or at the time of recurrence, the patient was either reexamined in the clinic or interviewed on the telephone. The patient was interviewed by telephone if he or she worked during the day and was unwilling to take time off from work or was from an outlying community and returning to the clinic presented a hardship. Fifty-three percent (n=50) of the patients were reexamined in the clinic, while 47% (n=44) were interviewed on the telephone. Sixty percent (30/50) of the no-vibration group and 46% (20/44) of the vibration group were examined in the clinic. The Dix-Hallpike maneuver was performed on patients reexamined in the clinic. Patients interviewed by telephone were instructed to initiate the head or body movement that had previously provoked the symptoms and to report the outcome.
The patients quantified their symptom intensity on a scale of 1 to 3 (mild, moderate, and severe) before the treatment procedure and at the time of follow-up. Results, defined as change, were categorized on a scale of 1 to 4 (cure, much better, better, and no change) on the basis of clinical examination or telephone interview at the time of follow-up. After 1 week, if the symptoms persisted, the canalith repositioning procedure was performed again or the patient was instructed to perform the Brandt-Daroff positioning exercises for 2 weeks. These results are not reported herein.
We also determined whether there was a recurrence of BPPV in the same or opposite ear during a 25-day to 5.25-year follow-up period. This was accomplished through reexamination of cured patients in the clinic or telephone interviews. If the patient was treated successfully with the canalith repositioning procedure with or without vibration and then later redeveloped BPPV symptoms, he or she was diagnosed as having recurrent BPPV.
Statistical analysis of the data was performed with Systat (Version 8; SPSS Corp, Chicago, Ill). Level of significance for all analyses was P<.05.
We identified 94 patients with BPPV involving the posterior semicircular canal, treated between July 1, 1991, and December 19, 1997. Of this population, 44 patients were treated with vibration and 50 patients, without vibration. Table 2 summarizes the demographic characteristics of the groups. There were no significant differences in age, sex, duration of symptoms before treatment, and intensity of symptoms before treatment. Before treatment, nearly half of the patients in each group described their symptoms as moderate in intensity, but not disabling.
Both variants of the canalith repositioning procedure were found to be effective in relieving symptoms of positional vertigo 1 week after treatment. There was no significant difference in the results between the 2 groups (Table 2).
Although treatment was effective, BPPV often recurred. The rate of recurrence of BPPV symptoms was determined on the basis of follow-up. Patients were excluded if they did not respond to the treatment procedure (n=14) or if the Brandt-Daroff exercises were started 1 week after treatment (n=10), leaving 70 patients available for analysis. Of the patients who did not respond to the treatment procedure, 57% (n=8) were treated with vibration, while 43% (n=6) were treated without vibration. Most often, Brandt-Daroff exercises were prescribed for patients who had a previous history of recurrence. Exclusion of this group may have resulted in a decrease in the estimate of the frequency of recurrence. Of the 70 patients, symptoms recurred in 47% (n=33) (Figure 2). The maximum length of follow-up was 5.25 years. The majority of recurrence of BPPV occurred within the first 2 years. Of these patients, 45% (n=15) were treated with vibration, while 55% (n=18) were treated without vibration. For this population, the homogeneity of the vibration and no-vibration groups was tested. The groups did not differ significantly.
To determine whether the rate of recurrence differed between the vibration and no-vibration groups and to account for patients entering the study at different times, the Kaplan-Meier product-limit method was used to estimate the survival function. Survival time was defined as the number of days from the day of treatment to the day the symptoms of BPPV recurred. The estimated survival function for the vibration and no-vibration groups was plotted (Figure 3). From the position of the traces, it appeared that patients who did not receive vibration had a longer time until recurrence than patients who received vibration. The log-rank test, however, showed that there was no significant difference in the survival distributions between the vibration and no-vibration groups, indicating that the time of recurrence was not different significantly between the 2 groups.
Once it was determined that there was no significant difference in the rate of recurrence between the 2 groups, the data were stratified to determine whether the age of the patient covaried with the time of recurrence of BPPV. Surprisingly, age did not correlate significantly with the time of recurrence as determined by the use of the Kaplan-Meier product-limit method and the log-rank test (P=.63).
We also examined whether history of recurrent BPPV significantly affected the short-term or long-term outcomes of the canalith repositioning procedure with and without vibration. On the basis of the patients' history before treatment, patients were classified into 2 categories: recurrent BPPV or 1 episode of BPPV. Recurrent BPPV was defined as sporadic periods of BPPV symptoms over time, while 1 episode of BPPV was defined as continuous BPPV symptoms over time. Of the 70 patients with data available for analysis, 65 patients had adequate documentation to determine if they had recurrent BPPV or 1 episode of BPPV. There was no significant difference in the history of recurrent BPPV between the vibration and no-vibration groups (Pearson χ2, P=.30). Of the 65 patients, 14 (39%) of the 36 patients in the vibration group and 8 (28%) of the 29 patients in the no-vibration group had a history of recurrent BPPV. At 1 week, there was no significant difference in the effectiveness of the treatment between patients with a history of recurrent BPPV and those with 1 episode of BPPV (Pearson χ2, P=.42). After treatment, there was no significant difference in the rate of recurrence of BPPV between patients with and without a history of recurrent BPPV (Pearson χ2, P=.25).
Our results suggest that, for the treatment of BPPV involving the posterior semicircular canal, vibration applied during the canalith repositioning procedure does not significantly affect short-term or long-term outcomes. There was no significant difference in the effectiveness of the treatment or the time to recurrence of BPPV between the group of patients receiving vibration and those receiving no vibration (Table 2; P<.05). Age and history of recurrent BPPV before treatment did not correlate with rate of recurrence.
The lack of a significant difference in the effectiveness of the treatment between the vibration and no-vibration groups suggests that vibration either is ineffective or has such a small effect that it is not detected with our sample size. These results differ from those of Li16 when he concluded that vibration was "critical to the success" of the canalith repositioning procedure in the treatment of BPPV.
We suggest that this difference resulted from the limited number of subjects within Li's no-vibration group, which could have led to the random occurrence of poor results. In support of this hypothesis is that the results Li reported for the canalith repositioning procedure performed without vibration differ greatly from those of other studies10,11,13,15,17 (Table 1). In these studies, the average success rate was 76% after 1 treatment session.10,11,13,15,17 This success rate is similar to the findings of our no-vibration group. Higher success rates are found when the effects of multiple treatments are reported.
Our source of vibration could not account for the differences found between our study and Li's,16 as we both used the same device. However, it is possible that other devices might produce better results, such as the handheld vibrator (Oster, model 126-11A; Sunbeam, Boca Raton, Fla) used by Epley.6,18
Modifications made to the canalith repositioning procedure may account for the differences in the results found between our study and that of Li.16 Li made gradual transitions between positions and used uninterrupted vibration over the mastoid of the involved side. In our study, we made rapid transitions between positions and stopped vibration during the transitions between positions. In our study, vibration was applied over the mastoid of the involved side once the position was assumed and was maintained for a minimum of 20 seconds or until the nystagmus stopped.
However, our 22% failure rate may be related to causes other than the method of vibration. The canalith repositioning procedure is designed to treat BPPV attributed to "canalithiasis." However, 2 other mechanisms have been proposed: cupulolithiasis and vestibulithiasis. Cupulolithiasis is defined as debris adherent to the cupula of the posterior semicircular canal. This mechanism has been supported by pathological findings.19,20Vestibulithiasis refers to free debris within the short arm of the posterior semicircular canal and possibly in the vestibule.19,21,22 Hypothetically, all 3 mechanisms appear feasible, but the canalith repositioning procedure is designed only to treat canalithiasis. Accordingly, the 22% failure rate could also be explained by postulating that the canalithiasis hypothesis is not always operant and that there may be other mechanisms that cause BPPV.
We were surprised to find a 47% recurrence rate when patients successfully treated by the canalith repositioning procedure were followed up for as long as 5.2 years. Previous studies have shown a 10% recurrence rate within 4 months after treatment.8 Our findings at 4 months were similar, with a 17% recurrence rate. This suggests that there is a cumulative increase in recurrence of BPPV over time. More extensive long-term studies determining the rate of recurrence need to be performed to confirm this observation. Given that BPPV commonly recurs, efforts to prevent recurrence seem indicated.
In summary, our results suggest that vibration applied during the canalith repositioning procedure does not significantly affect short-term or long-term outcomes. Recurrence was seen in 47% of the patients followed up for up to 5.25 years.
Accepted for publication December 3, 1999.
We thank Imke Janssen, PhD, for performing the statistical analysis.
Reprints: Timothy Carl Hain, MD, Northwestern Memorial Hospital, 675 N St Clair, 15-200, Chicago, IL 60611 (e-mail: firstname.lastname@example.org).