Effect of different maneuvers of repositioning on benign paroxysmal vertigo: a network meta-analysis

Abstract

Background

The effect of different maneuver repositioning on benign paroxysmal vertigo was explored by network meta-analysis.

Methods

The PubMed, Embase, Cochrane library, and web of science databases were searched for randomized controlled studies on the effect of manipulative repositioning on benign paroxysmal vertigo from database creation to September 1, 2024, Bayesian network meta-analysis, R4.4.1 was used for data analysis.

Results

Twenty-two articles (n = 2507) were included in this study, and the results of network meta-analysis suggested the following odds ratios (OR) relative to the control group (UT): Epley maneuver (EM) vs UT [OR = 7.9, 95% CI (3.21, 23.31)]; Gufoni maneuver (GFM) vs UT [OR = 5.1, 95% CI (1.25, 21.45)]; Gans Repositioning Maneuvers (GRM) vs UT [OR = 11, 95% CI (1.65, 83.85)]; Modified Epley maneuver (MEM) vs UT [OR = 9.83, 95% CI (1.55, 64.06)]; Semont’s maneuver (SM) vs UT [OR = 6.1, 95% CI (1.97, 18.46)]. The largest surface area under the cumulative ranking curve was for GRM (71.5%), followed by MEM (68%) and SEM (67.8%), and the worst was for UT (5.7%).

Conclusion

Based on our current findings, GRM, MEM and SEM are effective for BBPV symptoms, but due to the existence of study limitations, more high quality multicenter large sample randomized controlled studies are needed to testify to our conclusions.

Introduction

Benign Paroxysmal Positional Vertigo (BPPV), commonly known as "ear stone disease," is characterized by paroxysmal and recurrent vertigo triggered by changes in head position or body posture [1]. This condition is the most common peripheral vestibular disorder which is typically accompanied by nystagmus and autonomic symptoms. It presents with hearing loss, tinnitus, or neurological symptoms and signs, occurring periodically with episodes ranging from hours to years [2, 3]. Intervals between episodes are unpredictable and can span one to several years, with some cases remaining dormant for 10 to 20 years [4]. BPPV can affect individuals at any age but is most prevalent among those aged 50 to 70 years. Posterior canal BPPV is the most common subtype, accounting for 85–90% of cases, with a slightly higher incidence on the right side [5]. Horizontal canal BPPV accounts for 5–10% of cases, and anterior canal BPPV is rare, comprising about 2% [6]. The condition can resolve spontaneously but is prone to recurrence, with a recurrence rate of 20–30% and a 10-year recurrence rate of 50% [7, 8]. Research by von Brevern et al [9]. indicates that the incidence in females is approximately twice that of males. Independent risk factors for BPPV include hyperlipidemia, hypertension, migraine, and metabolic diseases such as type II diabetes, osteoporosis, gout, and vitamin D deficiency [10, 11]. Abnormal serum calcium metabolism leading to reduced bone density and osteoporosis may result in changes in otolith composition, causing otoliths to detach more easily from the utricular macula. Conditions such as tracheal intubation, head trauma, age-related degenerative changes, and diabetes can lead to vascular abnormalities in the small vessels supplying the inner ear, thereby impairing the peripheral vestibular system's ability to maintain balance [12].

Among the various treatment methods for BPPV, repositioning maneuvers are widely considered the most effective. These maneuvers involve specific head and body movements designed to relocate the dislodged otoliths back to the utricle, thus relieving the stimulation of the semicircular canal receptors. Commonly used repositioning maneuvers include the Epley maneuver, the Semont maneuver, and the Barbecue roll maneuver. While these methods have demonstrated efficacy in clinical practice, the relative effectiveness of each maneuver remains a topic of debate due to individual patient differences and the complex pathophysiology of BPPV [13, 14]. The Epley maneuver, one of the earliest and most widely applied repositioning techniques, is typically used for posterior canal BPPV. This method involves a series of head and body rotations to move the otoliths out of the posterior canal [15]. Studies have shown that the Epley maneuver effectively alleviates symptoms in the short term, though its long-term efficacy requires further investigation [16, 17]. The Semont maneuver, on the other hand, is a rapid and straightforward technique that achieves otolith repositioning through quick changes in body position. While effective for posterior canal BPPV, the rapid movements involved may cause discomfort or nausea in some patients [17]. The Barbecue roll maneuver, primarily used for horizontal canal BPPV, involves a sequence of rotational movements to expel the otoliths from the canal. Despite its simplicity, the effectiveness of this method varies depending on the type of BPPV [18], To comprehensively evaluate the efficacy of different repositioning maneuvers for BPPV, network meta-analysis (NMA) has become increasingly utilized in this field of research. NMA allows for the comparison of both directly and indirectly compared treatments, providing more robust evidence for clinical decision-making. The key advantage of NMA lies in its ability to integrate results from multiple studies, reducing bias and uncertainty associated with single studies and thus offering more reliable evidence.

Methods

Literature retrieval

The PubMed, Embase, Cochrane library, and Web of Science databases were searched for randomized controlled studies on the effects of manipulative repositioning on benign paroxysmal vertigo from the time the database was created until September 1, 2024, with the search term Benign Paroxysmal Positional Vertigo; Maneuver, and the specific search strategy is described in Supplementary Material Table S1.

Criteria for inclusion and exclusion

Inclusion criteria: The population included in this study was adults who met the diagnostic criteria for BBPV, and the interventions were Epley maneuver(EM); Usual treatment(UT) containing conventional medications; Semont's maneuver(SM); Brandt-Daroff(BD); Gans Repositioning Maneuvers(GRM); repeated Dix-Hallpike test(rDHt); Barbecue rotation maneuver(BRM); Gufoni maneuver(GFM); Self- Epley (SEM), and the primary outcome metric was remission rate (defined as disappearance of symptoms in BBPV). The main types of studies we included were randomized controlled studies.

Exclusion criteria: duplicates, animal experiments, protocols, conference abstracts, reviews, unavailable full text, articles with no available data.

Data extractions

The two authors (SNN and LMY) rigorously screened the literature based on the predefined inclusion and exclusion criteria. In case of any disagreement, they resolved it through discussion or sought the input of a third party (CW) for consultation and to reach a consensus. The extracted information from the included studies encompassed the following key details: first author, year of publication, country, sample size, gender, mean age, outcome.

Risk of bias assessment

Risk of bias is assessed by following the latest recommendations of the Cochrane Handbook Risk of Bias assessment tool [19], which consists of five main sections: bias arising from randomization, bias from deviations from established interventions, bias from missing outcome data, bias from outcome measures and outcome selective reporting bias. Against the literature, studies were rated as "Low risk of bias", "Some concerns", "High risk of bias ". The results are checked between the two researchers and if there is disagreement, a discussion is held to reach agreement, or a third person is consulted.

Data analysis

We conducted a Bayesian network meta-analysis using R4.3.2 software (R Foundation for Statistical Computing) with a priori fuzzy random effects models for multiple sets of trials. The combined estimates and probabilities of each treatment being the best were obtained by Markov chain Monte Carlo methods [20]. Model convergence was assessed by trajectory plots and Brooks-Gelman-Rubin plots dichotomous classification results were expressed as the posterior odds ratio or and its 95% confidence intervals (CI). We calculated the percentage of area under the cumulative ranking curve (SUCRA) to estimate the probability of optimal intervention. Network diagrams were drawn using STATA 15.0 with a pass-through macro command loaded. For the network diagram, each circle corresponds to a drug and the edges represent existing comparisons. The size of the circles is proportional to the number of patients included. Cumulative probability plots were drawn using the ggplot2 package.

Result

Literature screening

Figure 1 demonstrates the literature screening process, 2569 articles were retrieved by searching PubMed, Embase, web of science, Cochrane library, and 22 randomized controlled studies [21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42] were finally included by removing duplicates (n = 834), reading titles and abstracts for removal (n = 1697), and reading full text for removal (n = 16).

Fig. 1

Literature search flow chart

Basic characteristics and risk of bias

Table 1 shows the basic characteristics of 22 articles (n = 2507) with interventions containing EM: Epley maneuver; UT: BD: Brandt-Daroff; GRM: Gans Repositioning Maneuvers; rDHt: repeat Dix-Hallpike test; BRM: Barbecue rotation maneuver; GFM: Gufoni maneuver; SEM: Self-Epley. The included studies clearly accounted for the method of randomization used and were therefore evaluated as low risk, but some studies did not account for the blinding method used and were evaluated as unclear, the specific risk of bias is shown in Figs. 2 and 3.

Table 1 Literature characteristics table

Fig. 2

Risk bias of summary

Fig. 3

Risk bias of graph

Response rate

Twenty-two articles mentioned the response rate (defined as the disappearance of symptoms in BBPV). The network diagram (Fig. 4) suggests the formation of a closed loop; therefore, a local inconsistency test was used. The test results (Supplementary Materials Figure S1) indicated no statistical differences in the direct comparison, indirect comparison, and network comparison between SM vs BD; UT vs BD; MEM vs EM; SM vs EM; UT vs EM; UT vs MEM; and UT vs SM. The league table (Table 2) shows that EM vs UT [OR = 7.9, 95% CI (3.21, 23.31)]; GFM vs UT [OR = 5.1, 95% CI (1.25, 21.45)];GRM vs UT [OR = 11, 95% CI (1.65, 83.85)]; MEM vs UT [OR = 9.83, 95% CI (1.55, 64.06)]; SM vs UT [OR = 6.1, 95% CI (1.97, 18.46)]. The surface under the cumulative ranking curve was found to be greatest for GRM (71.5%), followed by MEM (68%), SEM (67.8%), and worst for UT (5.7%) (Fig. 5, Table 3).

Fig. 4

Network diagram; Each circle represents an intervention (EM: Epley maneuver; UT: Usual treatment; SM: Semont’s maneuver; BD: Brandt-Daroff; GRM: Gans Repositioning Maneuvers; rDHt: repeated Dix–Hallpike test; BRM: Barbecue rotation maneuver; GFM: Gufoni maneuver; SEM: Self-Epley)

Table 2 League table

Fig. 5

Cumulative probability ranking; The area under the curve (EM: Epley maneuver; UT: Usual treatment; SM: Semont’s maneuver; BD: Brandt-Daroff; GRM: Gans Repositioning Maneuvers; rDHt: repeated Dix–Hallpike test; BRM: Barbecue rotation maneuver; GFM: Gufoni maneuver; SEM: Self-Epley)

Table 3 Sucra rank

Publication bias

We used a funnel plot to evaluate the responds rate publication bias for the study, and the results (Supplementary Material Figure S2) suggest the possibility of a large publication bias.

Discussion

To our knowledge, this is the first time that a network meta-analysis has been used to explore the effects of different maneuvers on BBPV. The results of this study showed that EM, GFM, GRM, MEM, and SM all significantly improved the remission rate compared to the control group (UT). Specifically, GRM showed the greatest effect with an OR of 11 and a 95% confidence interval of (1.65, 83.85), which was significantly higher than the other treatments. These results provide a valuable reference for clinical treatment, suggesting that these treatments can significantly improve the clinical remission of patients, especially GRM, which is further supported by the SUCRA value, indicating that GRM (71.5%) is the most effective treatment option, with an efficacy ranking ahead of the other treatments. This was followed by MEM (68%) and SM (67.8%), which indicated that these three treatments were relatively better in terms of remission rates. Comparatively, UT had the lowest efficacy ranking (5.7%), further confirming its poor therapeutic outcome in this study. Thus, these SUCRA values not only support our conclusions reached through the OR values, but also provide a ranking guide for clinical treatment, helping physicians to choose the best treatment option based on efficacy.

Other findings related to GRM further support the findings of this study, in a study by Robert et al., results one week after a trial combining the Gans maneuver with post-maneuver restrictions showed a recovery rate of 80.2% after one intervention, 95.6% after two interventions, and 99% after three interventions [43]. Francesco Pandipusa et al. described the efficacy of GRM as the most comfortable procedure compared to Semont and Epley repositioning maneuvers [44]. Badawy WM et al. [45] studied the effects of GRM with and without post-maneuver instructions and found no differences; their sample size was very small. This indicates that GRM alone is an effective maneuver therapy without the need for post-maneuver instructions. Alia Saberi et al. [46] conducted a one-week follow-up study on 73 patients to determine the differences between GRM and ERM. As they mentioned in their study results, longer follow-up can be conducted in the future to examine the reversibility or recurrence of vertigo symptoms. Abir Omara et al. [47] compared GRM and EM in terms of postural stability and vertigo, finding both equally effective. GRM is equivalent to EM and can be used as a substitute due to its inclusion of neck extension and rotation. This will help raise awareness of this innovative BPPV treatment procedure. A recent systematic review [48] also indicated that the effectiveness of the Gans maneuver in medical research holds significant clinical importance. After implementing this technique, vertigo patients can achieve immediate relief without any supportive treatment, which not only has scientific value but also economic, social, and personal benefits. Moreover, for elderly and postural BPPV patients who experience neck stiffness and discomfort or other conditions, the GRM is equivalent to the Epley maneuver [49]. Therefore, the GRM can replace the Epley maneuver as it involves neck extension and rotation. This will help to raise awareness among the public and practitioners about this innovative BPPV treatment procedure.

EM and other repositioning methods can lead to some patients experiencing residual symptoms such as head dizziness, heaviness, unsteady walking, and neck stiffness after successful repositioning, with a high recurrence rate [4]. Later, a modified Epley maneuver was developed based on these methods, which effectively addressed the issues of residual symptoms and high recurrence rates. Because the modified Epley maneuver involves resting the patient's head on the bed (not in a suspended state), it is relatively easy to position the head, making the maneuver simple, labor-saving, self-treatable, and relatively safe. The SEM method is effective, safe, and easy to perform. After explanations and demonstrations by medical staff and watching repositioning videos, patients and their families usually master this method, allowing for self-repositioning treatment at home in case of recurrence [15]. This reduces over-reliance on medical personnel and alleviates psychological fear and anxiety in patients. The supplementary treatment provided by the SEM method addresses the difficulties faced by patients in remote areas and those with financial constraints, reducing the economic burden and fear of not being able to seek timely medical attention to some extent [50, 51].

The current study still has the following limitations, first: the frequency of interventions used in this included study, the time is not accounted for, second: for the inclusion of the population included all BBPV, which can lead to a greater heterogeneity in our study, and third: some of the studies have fewer indicators of the observed outcomes.

Conclusion

Based on our current findings, GRM, MEM and SEM are effective for BBPV symptoms, but due to the existence of study limitations, more high quality multicenter large sample randomized controlled studies are needed to testify to our conclusions.