1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Conflict of interest
  8. References


To evaluate the long-term efficacy of the Epley manoeuvre as a therapeutic procedure for posterior canal benign paroxysmal positional vertigo.


Randomised, double-blind, sham-controlled trial.


A multidisciplinary dizziness unit in a non-academic Hospital.


Forty-four patients with posterior canal benign paroxysmal positional vertigo (BPPV) with a duration of at least 1 month. Participants were randomised in two groups of 22 and treated with either the Epley manoeuvre or a sham manoeuvre and followed up for 1 year after treatment.

Main outcome measures

Conversion of a ‘positive’ Dix–Hallpike test to a ‘negative’ Dix–Hallpike test, impairments perceived by the dizziness assessed by the Dizziness Handicap Inventory (DHI).

Statistical analysis

Absolute and relative risks were computed, and Fisher's exact test was used to compare the treatments.


Six patients were lost to follow up (five in the sham group, one in the Epley group). The Epley procedure resulted in a treatment success in 20/22 patients (91%) after 12 months of follow-up, whereas the sham procedure had a positive effect in 10/22 patients (46%; P = 0.001). The DHI was significantly lower in the Epley group at all follow-up assessments (median scores 12 months 0 (0–51) versus 20 (0–76), P = 0.003).


The Epley manoeuvre provides long-term resolution of symptoms in patients with posterior canal BPPV.

Benign paroxysmal positional vertigo (BPPV) is the most common peripheral vestibular disorder. The incidence is estimated to be 64 cases per 100 000 population per year, the lifetime prevalence 2.4%.[1, 2] It is characterised by short-standing episodes of vertigo precipitated by a change in head position, the most common provoking movements being looking upwards, bending over and rolling over in bed. It is believed that dislodged otolithic debris inside one of the semicircular canals causes continuing movement of the endolymph, even after head movement has ceased.[3] This results in bending of the cupula, thus provoking vertigo. In a vast majority of cases, the posterior semicircular canal is affected.

The Dix–Hallpike manoeuvre confirms the diagnosis of posterior canal BPPV by eliciting vertigo as well as a typical nystagmus in the head-hanging position.

The most commonly utilised single treatment approach to posterior canal BPPV is the Epley manoeuvre. This manoeuvre involves a series of movements of the head and body in order to move the free-floating debris out of the posterior semicircular canal into the vestibule.[4] In the latest update of the Cochrane review on the effectiveness of the Epley manoeuvre, it is concluded that there is evidence that the Epley manoeuvre is a safe, effective treatment for posterior canal BPPV in the short term, but that there is insufficient evidence that the manoeuvre provides a long-term resolution of symptoms.[5]

The present study examines the long-term efficacy of the Epley manoeuvre in the treatment of posterior canal BPPV by comparing the difference in outcome between the Epley manoeuvre and a sham manoeuvre.

Patients and methods

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Conflict of interest
  8. References

Patients were recruited from the multidisciplinary dizziness clinic in Gelre Hospital in Apeldoorn. Those with an age over 18 years and a history of brief, position-related vertigo of at least 1 month were eligible for inclusion. An absolute requirement for inclusion in the study was a classic positive response to the Dix–Hallpike manoeuvre. This was defined as vertigo and upbeating and rotatory nystagmus beating towards the undermost ear in the head-hanging position on either the left or right side. In case of a bilateral positive DH manoeuvre, the side with the strongest nystagmus was seen as the affected side. Exclusion criteria were previous treatment with an Epley manoeuvre, cervical disc herniation and severe communication problems. The study was approved by the Medical Ethical Committee of the University Medical Center Utrecht.

Informed consent, randomisation

The senior investigators (T.B. and P.P.v.B.) obtained informed consent from each patient before baseline study assessment. After inclusion, the patients completed the Dutch version of the Dizziness Handicap Inventory (DHI). Then, they were randomly assigned with a 1 : 1 ratio to one of two treatment groups. Numbered, sealed envelopes containing the treatment group assignment were prepared before the start of study. With the use of a computer program, one of the investigators (J.C.) randomly assigned the numbers to either the Epley or the sham group. A technician (S.M. or P.O.) opened the next numbered envelope. Patients and outcome assessors were blind to allocation.


Those patients randomised to the Epley group were treated by an experienced technician (S.M. or P.O.) with an Epley manoeuvre. The Epley manoeuvre was performed as initially described by Epley, but none of the patients received premedication before treatment, nor did we use a mastoid vibrator.[4] The sham procedure was also performed by one of the technicians. It was similar to the Semont diagnostic manoeuvre: the patient was seated with the head turned 45° towards the unaffected ear; then, the patient was slowly tilted towards the affected ear (so that the patient lies on his side with the nose pointed upwards). Finally, the patient was slowly moved to the seated position again. After each Epley or sham manoeuvre, the Dix–Hallpike test was repeated. If necessary, a therapeutic manoeuvre (Epley as well as sham manoeuvre) was repeated up to two times.

After the therapeutic procedure, patients were advised to sleep propped up for 48 h and to avoid lying down on the affected side for 48 h in order to prevent repositioned particles from returning to the posterior canal.

Follow-up assessments were performed by a different technician from the one who performed the treatment, so that the assessor was blinded to which manoeuvre the patient had received. In each patient, the technician performed a Dix–Hallpike manoeuvre. Follow-up visits were planned at one, three, six and 12 months after the initial treatment. At each visit, the patient was asked to fill out the Dutch version of the DHI.[6] The DHI is a questionnaire which objectively quantifies the symptoms of dizziness. Its score ranges from 0 to 100 (higher scores imply more dizziness-related impairment).

Outcome and statistical analysis

The primary outcome was the proportion of patients with a negative Dix–Hallpike test at 12-month follow-up. Secondary outcomes were the Dix–Hallpike test results at the other follow-up moments, the DHI score and the occurrence of adverse events. Randomisation was checked on baseline characteristics. Differences between the groups at each follow-up assessment were tested with regard to the proportion of patients with a negative Dix–Hallpike test and median DHI score. Continuous variables were tested with the Student's t-test (normally distributed) or the Mann–Whitney U-test (skewed distribution), and nominal variables were tested with Fisher's exact test.

Apart from a complete case analysis, which uses only those patients who had complete follow-up assessments, we also used the last observation carried forward method (LOCF), in which the last result of the Dix–Hallpike test of a patient before dropping out was copied to all subsequent follow-up moments.

If there was any imbalance between the groups at baseline (P < 0.10), the relation of these variables with the outcome after 12 months was assessed to identify potential confounders. Potential confounding was adjusted for by multivariate logistic regression.

P values <0.05 were considered statistically significant.

Sample size calculations were based on a pilot study that we published in 2005.[7] Assuming a spontaneous resolution of 50%, a sample size of 23 patients per group was required to detect a difference of 40% in symptom resolution or negative DH test (two-sided significance level of 0.05, power 80%).


  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Conflict of interest
  8. References

Of more than 600 eligible patients, 44 patients (18 males, mean age 59.1, sd 13.0 years) with posterior canal BPPV (29 right sided, 15 left sided) consented to participate in this study between November 2006 and May 2011. Patients were randomised to the Epley (n = 22) or to the sham manoeuvre (n = 22). The Consort flow chart is shown in Fig. 1. Randomisation was successful on all baseline characteristics, but patients in the sham group were slightly older than those in the Epley group (62.5 years versus 55.7 years, P = 0.08) (Table 1), and patients in the Epley group had a somewhat lower median DHI score than patients in the sham group (23, range 8–66 versus 33, range 16 72, P = 0.08).

Table 1. Patient characteristics
 Epley (n = 22)Sham (n = 22)P value
  1. SD, standard deviation; BPPV, benign paroxysmal positional vertigo; DHI, Dizziness Handicap Inventory.

  2. a

    Compared with the Student's t-test.

  3. b

    Compared with the Fisher's exact test.

  4. c

    Compared with the Mann–Whitney U-test.

Mean age in years (sd)55.7 (12.8)62.5 (12.7)0.08a
Males10 (46%)8 (36%)0.76b
Presence of periodic dizziness9 (41%)9 (41%)1.0b
Time since first appearance dizziness in months [median (min–max)]10 (0–300)6 (0–660)0.99c
Mean duration of last episode in months (min–max)2.5 (0–24)2.3 (1–12)0.85c
Post-traumatic BPPV3 (14%)1 (5%)0.61b

Figure 1. Flow diagram.

Download figure to PowerPoint

During the trial, six patients were lost to follow-up: one in the Epley group, five in the sham group. Reasons were unrelated illnesses (n = 3), nausea/risk of falling (n = 2) and unknown (n = 1). Patients who were lost to follow up more often had a positive Dix–Hallpike manoeuvre (4/6, 67%) than patients who stayed in the study (10/38, 26%, P = 0.07). Because of this selective dropout, multiple imputation, which is the preferred method for imputing missing values, was not applicable.[8] We therefore chose to impute data with the last observation carried forward method.

The Epley procedure resulted in a long-term treatment success of 20/22 (91%, 95% CI 71–99%), whereas the sham procedure had an effect in 10/22 (46%; 95% CI 24–68%, P = 0.003), leading to a relative risk of the Epley manoeuvre for treatment success of 2.0 (95%, CI 1.2–2.6) (Table 2). The DHI was significantly lower in the Epley group at all follow-up visits (Table 3).

Table 2. Patients with negative Dix–Hallpike test result during follow-up
 Complete case analysisLast observation carried forward
EpleySham P a RR (95% CI)Epley (n = 22)Sham (n = 22) P a RR (95% CI)
  1. a

    Effect of treatment was tested using Fisher's exact test.

1 month20/21 (95%)9/20 (45%)<0.00012.1 (1.3–2.5)21 (96%)8 (36%)<0.0012.6 (1.5–3.1)
3 months19/21 (91%)9/18 (50%)0.0051.8 (1.1–2.4)20 (91%)9 (41%)<0.0012.2 (1.3–2.9)
6 months18/20 (90%)8/17 (47%)0.0041.9 (1.1–2.6)19 (86%)9 (41%)0.0022.1 (1.2–3.1)
12 months20/21 (95%)9/17 (53%)0.0021.8 (1.1–2.1)20 (91%)10 (46%)0.0012.0 (1.2–2.6)
Table 3. Dizziness Handicap Inventory (DHI) score over time
 Complete case analysisLast observation carried forward
Epley Sham  Epley (n = 22)Sham (n = 22) 
  1. a

    Differences in DHI score between groups were tested with the Mann–Whitney U-test.

DHI median score n   n   P a    P a
Baseline2223 (8–66)2233 (16–72)0.08n/an/a 
1-month follow-up225 (0–58)2024 (0–76)0.0055 (0–58)24 (0–76)0.003
3-month follow-up210 (0–40)1819 (0–72)0.0170 (0–40)20 (0–76)<0.001
6-month follow-up210 (0–38)1712 (0–80)0.191 (0–44)19 (0–80)0.009
12-month follow-up210 (0–46)1712 (0–64)0.190 (0–51)20 (0–76)0.003

Although age and DHI were not balanced between the two treatment groups, both variables were not related to the treatment outcome at 12 months (P value for age: 0.17, P for DHI: 0.86) and could therefore not confound the treatment effect.

We did not find any adverse events or complications of treatment. More specifically, there was no conversion of posterior canal BPPV into horizontal canal BPPV.


  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Conflict of interest
  8. References

Synopsis of key findings and clinical applicability

This study shows that the Epley manoeuvre is not only an effective treatment for posterior canal BPPV in the short term, but also in the long term. After 12 months, symptoms had resolved in 91% of patients treated with the Epley manoeuvre compared with 46% of patients treated with the sham manoeuvre.

The treatment results in the Epley group were relatively straightforward: 21 of 22 patients (95%) were BPPV free directly after treatment, and this result consisted during all follow-up visits. Of these 21 patients, one developed BPPV at the contralateral side and two had a recurrence which spontaneously resolved before the end of the follow-up period. One patient did not respond to treatment and remained BPPV positive at all visits.

The results in the sham group were diverse: two patients withdrew from the study directly after randomisation, leaving 20 patients in this group. Ten patients (50%) initially responded to the sham treatment of the affected side. Of these, four remained negative at all follow-up visits, two had a recurrence on the affected side which spontaneously resolved during follow-up, and three developed a recurrence which did not spontaneously resolve. One had a recurrence which resolved, recurred again and had resolved at the final visit at 12 months. Ten patients initially did not respond to the applied sham manoeuvre; three remained positive at all follow-up visits, two remained positive and also developed BPPV at the contralateral side, and three were initially not responsive but the BPPV spontaneously resolved during follow-up. One had a spontaneous resolution after initially not responding, but at the final visit the BPPV had recurred, and in another one the initially unresponsive BPPV at the affected side had resolved, but BPPV had developed at the contralateral side. The difference in recurrence rate of BPPV in patients who underwent an Epley manoeuvre and a sham manoeuvre suggests that treatment of BPPV with the Epley manoeuvre lowers the risk of recurrence of BPPV.

Comparison with other studies

Our results are in line with those from a pilot study (comparing the Epley manoeuvre with no treatment) that we published some years ago.[7] Epley's original study demonstrated a success rate of 97.7%.[4] Although subsequent clinical trials reported somewhat lower success rates (Lynn 89%, von Brevern 80%), the overall picture is one of the very effective therapeutic procedures.[9, 10] A Cochrane systematic review included five randomised controlled trials of the Epley manoeuvre versus placebo (sham treatment) or no treatment. Pooled data (including 292 adults with BPPV) showed a statistically significant effect in favour of the Epley manoeuvre over controls.[5] However, the follow-up period in these trials was brief, varying from 24 h to 1 month.[9-13]

In the literature, variable recurrence rates have been reported. Nunez et al. calculated a recurrence rate of ≈15% per year among patients who had initially reported resolutions of symptoms after canalith repositioning manoeuvre, whereas Beynon et al. reported a recurrence rate of BPPV after treatment with the Epley manoeuvre as high as 44% at 30 weeks or longer follow-up.[14, 15]

Strengths and limitations of the study

An important issue is the natural resolution of BPPV. We know that spontaneous resolution of symptoms of BPPV is common, although percentages in the literature vary greatly. In a pilot study comparing the Epley manoeuvre with no treatment, we found 50% spontaneous resolution in the untreated group after 6 months (compared with 90% resolution in the Epley group).[7] One study found a percentage as high as 85% of symptom resolution in the control group after 3 months.[16] However, in this particular study, most patients had symptoms for only 2 weeks or less. In these patients, one would expect spontaneous resolution of symptoms to occur easily. This is why we only included those patients with symptoms for at least 1 month.

Some other points regarding the present study need to be considered. First, the number of patients included in our study was rather small (considering an inclusion period of 4.5 years). It is well known that sham procedures hinder participation rates in trials. Indeed many patients who were asked to participate did not wish to undergo a sham procedure, while an effective treatment was available. Alternative study designs have been proposed since the start of our study, such as the stepped-wedge design, which we should, with the current knowledge, have considered in the study presented here.[17]

Second, the baseline characteristics were not comparable between the Epley and the sham group, the mean age in the Epley group being somewhat lower as compared to that in the sham group. However, we did not have to perform multivariate regression analyses because the imbalanced variables were not related to the outcome. It is unlikely that this has influenced the results, as it is well known that age has no impact on the success rate of the Epley manoeuvre.[18]

Considering the total number of included patients, the number of patients lost to follow-up was relatively high (n = 6), especially in the sham group (n = 5). Complete case analysis is known to produce biased results, and because our missing data were not random, we used the LOCF method to impute the missing outcomes. When comparing complete case analysis with the LOCF method, both analyses show a beneficial effect of the Epley manoeuvre in the long term, which strengthens our conclusion that the Epley manoeuvre has a long-term beneficial effect in patients with BPPV.

  • The Epley manoeuvre is a highly effective treatment for posterior canal BPPV in both the short and long term.
  • The Epley manoeuvre probably reduces the BPPV recurrence rate.


  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Conflict of interest
  8. References

We would like to thank Sylvia Masius-Olthof and Peter Oostenbrink, technicians, and Carla Colijn, research nurse for their contribution to this study.


  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. Conflict of interest
  8. References
  • 1
    Froehling D.A., Silverstein M.D., Mohr D.N. et al. (1991) Benign positional vertigo: incidence and prognosis in a population based study in Olmsted County, Minnesota. Mayo Clin. Proc. 66, 596601
  • 2
    von Brevern M., Radtke A., Lezius F. et al. (2007) Epidemiology of benign paroxysmal positional vertigo: a population based study. J. Neurol. Neurosurg. Psychiatry 78, 710715
  • 3
    Parnes L.S. & McClure J.A. (1992) Free-floating endolymph particles: a new operative finding during posterior semicircular canal occlusion. Laryngoscope 102, 988992
  • 4
    Epley J.M. (1992) The canalith repositioning procedure: for treatment of benign paroxysmal positional vertigo. Otolaryngol. Head Neck Surg. 107, 399404
  • 5
    Hilton M.P. & Pinder D.K. (2012) The Epley (canalith repositioning) manoeuvre for benign paroxysmal positional vertigo. Cochrane Database Syst. Rev. (2), CD003162
  • 6
    Vereeck L., Truijen S., Wuyts F.L. et al. (2006) Test-retest reliability of the Dutch version of the Dizziness Handicap Inventory. B-ENT 2, 7580
  • 7
    Richard W., Bruintjes Tj.D., Oostenbrink P. et al. (2005) Efficacy of the Epley maneuver for posterior canal BPPV: a long-term, controlled study of 81 patients. Ear Nose Throat J. 84, 2225
  • 8
    Greenland S. & Finkle W.D. (1995) A critical look at methods for handling missing covariates in epidemiologic regression analyses. Am. J. Epidemiol. 142, 12551264
  • 9
    von Brevern M., Seelig T., Radtke A. et al. (2006) Short-term efficacy of Epley's manoeuvre: a double-blind randomised trial. J. Neurol. Neurosurg. Psychiatry 77, 980982
  • 10
    Lynn S., Pool A., Rose D. et al. (1995) Randomized controlled trial; of the canalith repositioning procedure. Otolaryngol. Head Neck Surg. 113, 712720
  • 11
    Froehling D.A., Bowen J.M., Mohr D.N. et al. (2000) The canalith repositioning procedure for the treatment of benign paroxysmal positional vertigo: a randomized controlled trial. Mayo Clin. Proc. 75, 695700
  • 12
    Yimtae K., Srirompotong S., Srirompotong S. et al. (2003) A randomized trial of the canalith repositioning procedure. Laryngoscope 113, 828832
  • 13
    Munoz J.E., Miklea J.T., Howard M. et al. (2007) Canalith repositioning maneuver for benign positional vertigo. Can. Fam. Physician 53, 10481053
  • 14
    Nunez R.A., Cass S. & Furman J.M. (2000) Short- and long-term outcomes of canalith repositioning for benign paroxysmal positional vertigo. Otolaryngol. Head Neck Surg. 122, 647652
  • 15
    Beynon G.J., Baguley D.M. & da Cruz M.J. (2000) Recurrence of symptoms following treatment of posterior semicircular canal benign paroxysmal positional vertigo with a particle repositioning manoeuvre. J. Otolaryngol. 29, 26
  • 16
    Asawavichianginda S., Ispradit P., Snidvongs K. et al. (2000) Canalith repositioning for benign paroxysmal positional vertigo: a randomized, controlled trial. Ear Nose Throat J. 79, 732737
  • 17
    Brown C.A. & Lilford R.J. (2006) The stepped wedge trial design: a systematic review. BMC Med. Res. Methodol. 8, 54
  • 18
    Macias J.D., Lambert K.M., Massingale S. et al. (2000) Variables affecting treatment in benign paroxysmal positional vertigo. Laryngoscope 110, 19211924