Specific and individualized instructions improve the efficacy of booklet-based vestibular rehabilitation at home – a randomized controlled trial (RCT)

2.1 Participants

Seventy-four adult study participants with chronic dizziness and balance disorders (symptoms for more than three months) were recruited for the study at the outpatient clinic of the German Center for Vertigo and Balance Disorders (DSGZ) at the Ludwig-Maximilians University of Munich. Potential participants were identified and asked to participate in the study after outpatient work-up. The recruiting period started in 2016 and ended in 2020. Patients with chronic dizziness, vertigo, and balance problems (>3 month) due to peripheral (uni- or bilateral hypofunction) or central vestibular (brainstem, cerebellum) disorders were included.

Exclusion criteria were: (1) episodic vertigo syndromes such as vestibular migraine, Meniere’s disease or benign paroxysmal positional vertigo; (2) recent changes (last 3 month) in the medication or planned changes during the two-month of individual study duration; (3) patients with orthopedic or neurological diseases that affect balance; (4) patients with contraindications to VRT (severe pain, severe cervical syndrome); (5) patients unable to follow verbal and written instructions or not able to give written informed consent; (6) patients who, unrelated to the study, received physiotherapy or occupational therapy interventions to treat dizziness and balance.

All study participants (n = 74) had the initial assessment (T0) and were randomized into the intervention group (n = 37) and the control group (n = 37). The randomization was done by the online software tool Randoulette (provided by the Institute for Medical Information Processing, Biometry and Epidemiology –IBE at the LMU Munich; www.en.ibe.med.uni-muenchen.de/services/studienrandomisierung/index.html). Block randomization (block sizes of 2, 4 and 6) ensured a 1:1 ratio in terms of size of the intervention and control groups, peripheral and central diseases, sex and age. Study participants and investigators were blinded to group allocation. The physiotherapist was not blinded. After four weeks of training, the post-intervention assessment (T1) was done and the follow-up examinations (T2) were executed after another four weeks (Fig. 2). Participants were advised to continue exercising during the follow-up period. The follow-up could be carried out in a total of 68 study participants (Fig. 1). The dropouts were included in the descriptive and comparative statistics of the participant characteristics at T0.

OPEN IN VIEWER

Fig. 1

Flow chart of the study.

OPEN IN VIEWER

Fig. 2

Course of the study.

The study protocol has been approved by the Ethics Committee of the University of Munich (370-15). All procedures were in accordance with the Helsinki declaration and all study participants gave their written signed informed consent. The study was registered with the German Register of Clinical Studies (DRKS ID: DRKS00009016).

2.3 Outcome measures

The study participants were assessed with a battery of clinical and objective balance measures, as well as self-assessment instruments. The Dizziness Handicap Inventory (DHI-G) total score was the primary outcome. Secondary outcomes were the Visual Analog Scale for dizziness and balance disorders (VAS), the Activity-specific Balance Confidence scale (ABC-G), the Vestibular Activity and Participation Questionnaire (VAP-G), instrumented static posturography, the Timed-up-and-Go test (TUG), the Falls Efficacy Scale (FES-I), the therapy diary, the Problematic Experiences of Therapy Scale (PETS), and the accuracy scale.

All outcomes were assessed on admission to the study (T0), after four weeks of training (T1) and at follow-up after another four weeks (T2). Assessments were carried out by a blinded medical-technical assistant from the center (DSGZ). The study participants from both groups were recommended to continue exercising after the actual intervention (Fig. 2).

The Dizziness Handicap Inventory (DHI -G) was designed to assess the self-perceived handicap and quality of life in patients with dizziness and balance disorders. Dizziness is a major factor that reduces opportunities for participation and thus has an impact on quality of life [33]. The DHI-G has 25 items with three choices each. A “yes” response is awarded 4 points, a “sometimes” 2 points and “no” 0 points. The total score has a maximum of 100 points. The higher the score the greater the impairment caused by dizziness and imbalance. The DHI-G has three subscales: emotional (e), functional (f), physical (p). The DHI-G is a reliable instrument to record the specific impairments of those affected [27] and is widely used in the literature [12, 17, 18, 26].

The Visual Analog Scale for dizziness and balance disorders (VAS) is a graphic rating scale for measuring the subjectively perceived vertigo. The VAS was evaluated for four situations: general current condition, sitting without movement, during movement (e.g., walking), and during one-legged stance. The participants rated the intensity of vertigo symptoms using a slider on a 10 cm long line with the endpoints “no dizziness” and “unbearable dizziness” [12, 20]. For analysis a total score using the mean of all 4 situations for each participant was used.

The Timed-up-and-Go test (TUG) is a functional test to evaluate postural control, locomotion and risk of falling. The TUG measures the time taken by the patients to get up from the seat (standard armchair), walk a three-meter-long distance as quickly as possible (without running), turn around, go back to the chair and sit down. The time is measured in seconds. Values up to 20 seconds relate to independent mobility. Values between 20 to 30 seconds are associated with mild impairment of mobility and values >30 seconds indicate pronounced mobility impairment [37, 48].

Static posturography was used to objectively assess equilibrium on a stabilometer platform (Kistler 9261A, Kistler Group, Winterthur, Switzerland). Body sway path in X, Y, Z was measured under 10 different conditions for 30 seconds each [25]. The 10 different conditions are (1) standing on firm surface with eyes open (EO), (2) standing on firm surface with eyes closed (EC), (3) standing on firm surface with head reclination EO, (4) standing on firm surface with head reclination EC, (5) standing on foam with EO, (6) standing on foam with EC, (7) standing on foam with head reclination EO, (8) standing on foam with head reclination EC, (9) standing on foam in tandem stance EO and (10) standing on foam in tandem stance EC. For the study we analyzed the four most informative conditions in dizzy patients, i.e., conditions 1, 2, 5, and 6.

The Activity-specific balance confidence scale (ABC-G) is a questionnaire for recording fall-related self-efficacy. It contains 16 items covering activities from simple to complex. The individual items range from 0 % (not at all confident) to 100 % (absolute confident) [12, 38, 41].

The Falls Efficacy Scale-International (FES-I) is based on 16items, which are rated according to “how concerned you are about the possibility of falling”, using a four-point scale: not at all (1), somewhat (2), fairly (3), and very concerned (4). The total score ranges from 16 to 64 points. Higher scores indicate less fall-related self-efficacy. The lowest value of 16 means not any fear of falling [10, 31, 51].

The Vestibular Activity and Participation Questionnaire (VAP-G) is a 34-item questionnaire. We used the short form with 20-items and two subscales [35]. Scale one examines activities that trigger dizziness and impaired balance and scale two evaluates the consequences of this. The 5-point scale is defined as follows: (0) no difficulty, (1) mild, (2) moderate, (3) severe, (4) unable to do, and (5) not applicable. The higher the score, the greater the effect of balance and dizziness on an individual’s activities and participation [1, 34].

The Problematic Experiences of Therapy Scale (PETS) is a self-report measure that assesses the extent to prevent from carrying out an intervention by plausible reasons. It can be used to assess participants’ perceived reasons for non-adherence to a home-based rehabilitation therapy [24]. It is a 12-item questionnaire with three sections: problems due to symptoms, problems due to uncertainty or doubt about therapy, or practical problems. Participants rate the questions as “strongly agree”, “tend to agree”, “not sure”, “tend to disagree”, and “disagree”.

The Accuracy scale for exercise performance is a self-report measure. Each exercise can be evaluated in “exercise modified” (yes/no) and “is the exercise execution correct”. It is rated on a 4-point Likert Scale as follows: The exercise performance is correct. (1) strongly agree, (2) agree, (3) disagree, (4) strongly disagree. An independent blinded physiotherapist assessed the participants’ individual exercises using video recordings.

2.4 Intervention

To minimize systematic errors, all study participants were assessed and instructed by the same physiotherapist, an expert in VRT and a specialist therapist for dizziness and balance disorders (SK). The therapist kept the time with the study participants’ constant at 60 minutes in both groups. Study participants in both trial arms received a brochure for VRT at home (see supplementary material). The exercise selection was based on documented evidence for vestibular rehabilitation [11, 15, 29, 47]. The booklet includes 16 illustrated exercises for adaptation, habituation and substitution in vestibular dysfunction. The following eight weeks (intervention and follow-up), all study participants were asked to exercise at home for 20 minutes twice a day using the flyer. In order to monitor the therapy, they kept a standardized diary.

Intervention group: Study participants received a 60 min session with the experienced physiotherapist at T0. This first session included a physiotherapeutic assessment with a detailed investigation of the individual symptoms and the current activity level. Based on this assessment, appropriate exercises from the flyer were selected by the therapist and adapted to the individual needs of the study participant. The selection was based on the study participant’s impairments and adapted to the current level of performance by using the appropriate starting position (sitting, standing, narrow stand, tandem stand, balance pad), speed, amplitude, frequency, and duration. The modification of vision (eyes open, eyes closed) was also adapted individually. Participants were instructed to progress to more difficult exercises when possible. In addition, each exercise in the intervention group was demonstrated by the therapist, practiced with the study participant and explained in detail. The training principles and the continuous adaptation of the exercises were explained by the therapist. In addition, the study participants were given general advice on dizziness and balance, safety issues, and on avoiding falls.

Control group: Study participants in the control group also received a 60 min session with the same experienced physiotherapist after receiving the booklet at T0. The study participant’s symptoms and activity level were assessed as described above. Participants were instructed to do training for 20 min twice each day. However, no individual exercises were selected from the training block for the study participants, no individual program was provided and no further instructions were given. Training performance, intensity, duration or execution of the exercises were not explained further by the therapist. Study participants were only given general advice on dizziness and balance, safety instructions, and general recommendations for an active lifestyle and aids to prevent falls.

2.5 Data analysis

Data were analyzed using SPSS (IBM SPSS Statistics, Version 26). Data are reported as mean±SD. The effects of each dependent variable was analyzed using an analysis of variance for repeated measures (repeated measure ANOVA) with group (intervention vs. control) and time point (T0, T1 and T2) as factors and Bonferroni post hoc analysis. The homogeneity of variance was confirmed using the Levene test (p-value greater than 0.05). The requirement for sphericity was checked using the Mauchly test and, in the case of inhomogeneous factor levels, subjected to the Greenhouse-Geisser correction procedure. The evaluation method for the study was carried out as per protocol analysis. Results were considered significant if p < 0.05.

3.1 Participant characteristics

Table 1 shows baseline characteristics of study participants in the two groups. At baseline, the groups were comparable with respect to all assessed characteristics. The majority of study participants were between 60 and 79 years old. Only one study participant required professional assistance. Most study participants were able to perform daily life activities without any support (n = 41) or only required minor support from family and friends (n = 32).

Severe adverse events of the intervention did not appear (falls, fractures, hospitalizations). Study participants were informed before the start of the intervention that symptoms (dizziness) can worsen during therapy. ANOVA results are summarized in Table 2.

At baseline, DHI-G total revealed equal dizziness-related limitations of daily activities in the intervention and control group (p = 0.796). Both groups reported an improved DHI-G total score after 4 weeks of VRT (T1, p < 0.001), and further improvement at follow-up (T2, p < 0.001). Home-based exercises with instructions in the intervention group yielded greater improvements in DHI-G total score at T1 (p = 0.045) and T2 (p = 0.030) compared to controls (Fig. 3A). These general effects were differently manifested at the level of DHI subscores. With respect to DHI-G (p) reflecting physical activity and symptom triggers, both groups showed equal improvements at T1 and T2 (both p < 0.001) (Fig. 3B). With respect to DHI-G (e) and (f) reflecting emotional and functional aspects respectively, both groups showed reduced subscores at T1 and T2 (both p < 0.001), with greater improvement in the intervention group at T2 (DHI-G (e): p = 0.016; DHI-G (f): p = 0.042) (Fig. 3CD).

OPEN IN VIEWER

Fig. 3

Results for the DHI-G. Mean±SD are shown. A: DHI-G total score; B: DHI-G subscore (p); C: DHI-G subscore (e); D: DHI-G subscore (f). Abbreviations: DHI-G - Dizziness Handicap Inventory, p - physical, e - emotional, f –functional;; T0 –baseline; T1 –after 4 weeks of training; T2 –at follow-up (8 weeks after T0). Significant differences between groups are marked by an asterisk (*).

At baseline, subjectively perceived vertigo as assessed by the VAS was comparable between the intervention and control group (p = 0.082). Both groups reported a reduced intensity of symptoms after 4 weeks of VRT (T1, p < 0.001) and a further improvement at follow-up (T2, p < 0.001). Home-based exercises with instructions in the intervention group yielded greater reductions in VAS at T1 (p = 0.002) and T2 (p < 0.001) (Fig. 4A).

OPEN IN VIEWER

Fig. 4

Results for VAS, ABC-G, TUG, FES-I, VAP-G, and PETS. Mean±SD are shown. A: VAS total, B: ABC-G, C: TUG, D: FES-I, E: VAP-G, F: PETS. Abbreviations: ABC-G - Activities-Specific Balance Confidence Scale; FES-I - Falls Efficacy Scale International-G; PETS - Problematic Experiences of Therapy Scale; TUG - Timed up & go test; VAP-G - Vestibular Activity and Participation Questionnaire; VAS: Visual Analog Scale for dizziness and balance disorders; T0 –baseline; T1 –after 4 weeks of training; T2 –at follow-up (8 weeks after T0). Significant differences between groups are marked by an asterisk (*).

Fall-related self-efficacy levels as assessed by the ABC-G were comparable between groups at baseline (p = 0.131). Both groups reported reduced ABC-G at T1 (p < 0.001) and further improvement at follow-up (p < 0.001). The intervention group showed greater reductions in ABC-G at T1 (p = 0.005) and T2 (p = 0.027) (Fig. 4B).

At baseline the TUG performance was comparable between both groups (p = 0.092). Both groups showed improved TUG performance at T1 (p = 0.004) and a further improvement at T2 (p < 0.001). Home exercise with instructions in the intervention group resulted in greater improvements in the TUG performance at T1 (p = 0.020) and T2 (p = 0.008) (Fig. 4C).

At baseline, the FES-I was comparable between both groups (p = 0.173). Intervention and control group reported reduced FES-I T1 (p < 0.001) and a further reduction at follow-up (p < 0.001). This reduction was more pronounced in the intervention group at T1 (p = 0.007) and T2 (p = 0.013) (Fig. 4D).

At baseline, the VAP-G was not different between groups (p = 0.408). Both groups reported a reduced total VAP-G score at T1 (p = 0.001) and further improvement at T2 (p < 0.001). The home exercises resulted in a greater reduction in total VAP-G score at T2 (p = 0.013) in the intervention group (Fig. 4E).

In the PETS, participants documented reasons that hindered them from performing the exercises. In general, participants in the intervention group that were provided with individualized instructions for their home exercises reported less reasons that kept them from performing exercises at T1 (p = 0.001) and T2 (p = 0.020) (Fig. 4F). In the control group, the most prevailing reasons for missing exercises were associated to limited resources of time (Table 3).

The instrumented static posturography showed no difference between the intervention and control group in any of the examined stance conditions (Fig. 5).

OPEN IN VIEWER

Fig. 5

Results of static posturography (sway path). Mean±SD are shown. A: standing on firm surface with eyes open, B: standing on firm surface with eyes closed, C: standing on foam with eyes open, D: and standing on foam with eyes closed. Abbreviations: T0 –baseline; T1 –after 4 weeks of training; T2 –at follow-up (8 weeks after T0).

The evaluation of the therapy protocols showed a high training compliance in both groups. More than 80% of the study participants performed the exercises daily. In less than 10%, training took place only once a day but was mostly supplemented by additional sports activities. There was no significant difference between the two groups with respect to training intensity.

Compared to controls, individualized instructions for home-based VRT yielded more pronounced training improvements in the primary outcome measures, i.e., DHI-G. Analogous effects were found for the DHI-G subscores (p) and (e) and most secondary outcomes with the exception of the VAP-G and measures of standing balance from posturography.

This study examined the effect of an exercise therapy intervention.

4.1 Vestibular rehabilitation is an effective treatment

Previous research has proven that vestibular training is in many cases the most effective treatment option for dizzy patients [14]. The evidence for VRT effects is in particular high for patients with peripheral vestibulopathy [24]. Patients with central vestibular disorders also benefit from VRT [9, 40]. However, there is still insufficient evidence to distinguish between different forms of vestibular rehabilitation. Despite the strong evidence for vestibular rehabilitation a large proportion of dizzy patients report difficulty in getting access to VRT programs [30, 44, 50].

4.2 VRT as a home exercise program

Advice to do exercises at home is often given. Instructions are provided offline (printed booklet, flyer) or online (app, website) [46, 50, 52]. There is evidence that home exercise programs can improve symptoms in dizzy patients [2, 7, 8, 13, 21, 22, 28, 42, 43, 50, 52]. In our study, both groups of study participants improved after four and after eight weeks of booklet-based home training. This is consistent with previous findings [2, 5, 6, 9]. Home exercise programs in general have a positive effect on patients with chronic dizziness [15, 21, 22, 28, 46, 50, 52]. Self-training at home is also done to compensate for the poor access to hands-on vestibular rehabilitation [5, 30, 49]. However, many patients in home-based programs receive training material without individualized instructions. Corresponding to existing literature our results showed the effectiveness of a booklet-based vestibular training at home in reducing dizziness [45].

4.3 Individualized patient-specific instructions

Our study was able to show that specific and individualized physiotherapeutic support is beneficial for the outcome of VRT training at home. There was a significant reduction in dizziness symptoms, demonstrated as an improvement of the DHI-G score. The physiotherapist addressed the specific problems of participants and selected exercises based on standard assessments. Therapy complexity, starting positions, and specific exercises could be tailored to the individual subject. With the precise instructions participants are able to adapt the exercises in intensity, speed, and starting positions by themselves. The findings are consistent with existing literature on tailored VRT [13, 43]. In contrast to van Vugt et al. [46], we were able to show that a specific physiotherapeutic intervention with individualized instructions has a significant additional effect on reducing perceived symptoms and associated limitations. Our results suggest that participants benefit from accurate information. We assume that a tailored adaption of training to the patient’s needs can enhance compensation of the vestibular deficits. Patients guided by specific physiotherapy have a higher chance of reducing their dizziness on the VAS by at least 50%. A decrease of the intensity of dizziness might lead to a reduced use of health system services and thereby contribute to efficient and cost-saving patient care. Thus, we highly emphasize the recommendation of therapy adapted to the individual needs of patients.

4.4 Limitations of the study

The study has some limitations. Almost 90% of our cohort suffered from peripheral vestibular disorders, only about 10% (n = 8) had central vestibular syndromes. It is thus not clear if the findings can be generalized to all dizzy patients. Further, our sample from a tertiary dizziness center might not be representative for the situation in the practice of physiotherapists and general practitioners. However, the highly selected chronic patients at our center should be more difficult to treat. Other patients will therefore most likely also benefit from tailored training programs. For safety reasons, we decided not to provide standardized exercises with eyes closed for home use. Although previous work has shown that after vestibular loss, exercises on an unstable surface with and without eyes closed are crucial for compensation [6, 16, 23] we had to minimize the risk of falls in the home setting. Patients were encouraged, however, to do the training at the limit and as difficult as tolerated. The present study used standardized instruments for assessment. In particular the primary outcome measure DHI-G is frequently used in the literature. However, only static but not dynamic balance has been measured. Dynamic conditions are relevant as many patients report on difficulties during walking. However, we believe that the general limitation of the study participants was well covered by the available measures. In analogy to most therapy studies, we were not able to blind the therapist. We, however, do not think that this may have impacted the study outcomes, since all analysis was performed by investigator not involved in the intervention.