Graphic representation of the burden of suffering in dizziness patients

This research received no grants from any funding agency in the public, commercial or non-profit sectors. The study is part of a cross-sectional research project investigating patients with dizziness [4]. It was approved by the ethics committee of the Canton of Zurich, Switzerland. All participants gave their written informed consent prior to data collection. From August 2010 to August 2011, we asked patients between the ages of 18 and 65 years, inclusive, who had been referred to the Interdisciplinary Centre for Vertigo and Balance Disorders at University Hospital Zurich to participate in the study. They received questionnaires by mail and were asked to return them at the time of their clinical consultation. Patients were excluded from the study if the PRISM tool was missing.

The Pictorial Representation of Illness and Self Measure (PRISM) is a very brief and simple visualization tool that patients can use to rate their burden of suffering [20]. With the PRISM, patients depict the distance between their illness and themselves by positioning a disc within a rectangular field in which a circle in one corner represents the person’s self [22]. The distance between the illness and the self (the distance between the centres of the two discs) is called self-illness-separation (SIS), which is a direct measure of a patient’s perception of the controllability of their illness, and an inverse measure of its intrusiveness; in other words, the smaller the SIS, the greater the self-perceived intrusiveness of disease [26]. In this way, it is also inversely correlated with self-perceived of suffering, in that a small SIS indicates high levels of suffering. The PRISM has been successfully validated and used as an outcome measure for a variety of clinically-distinct diseases and health states, including lung diseases [26],[32],[33], rheumatoid arthritis [20],[26],[34],[35], systemic lupus erythematosus [24]-[26], multiple sclerosis [22], grief and trauma [22],[27],[28],[33], pain [23],[36], shared decision making [37], risk perception [38], cancer [33],[39],[40], dermatological diseases [29],[30],[33],[41], fertility problems [33], and diabetes [42]. For our study, the paper and pencil version was used. The measurement derived from the PRISM is the self-illness- separation (SIS), which ranges from 0 to 90 mm (the distance between the patient’s pencil mark and the centre of the self-circle). In terms of interpretation, two SIS ranges were distinguished (a mark within the self-circle, and a mark outside the self-circle) [26]. Previous studies have identified no disadvantages of the paper and pencil versus the original version, which uses a disc instead of a pencil mark [23],[43],[44]. The test and its instructions are shown in Figure 1.

The Dizziness Handicap Inventory (DHI, German version) is a disease-specific, self-rating questionnaire that assesses the patient’s severity of dizziness [45],[46]. It consists of 25 items, each having 3 response options (yes = 4, sometimes = 2, no = 0) and total scores ranging from 0 to 100. Three sub-scales measure the functional (DHIF, 9 items; e.g., Does your problem interfere with your job or household responsibilities?), physical (DHIP, 7 items; e.g., Does walking down the aisle of a supermarket increase your problems?) and emotional (DHIE, 9 items; e.g., Because of your problem, do you feel frustrated?) impact of dizziness. The German version has demonstrated good internal consistency (α = 0.72 to 0.89) and reliability (test-retest reliability r = 0.92 to 0.97) and thus is recommended as a measure of symptom severity in patients with dizziness [11],[46].

The Hospital Anxiety and Depression Scale (HADS, German version) was used to assess symptoms of emotional distress [47],[48]. Each of the 14 items is rated on a scale from 0 to 3, resulting in a summation score between 0 and 42 [49]. The scale has been shown to be an effective measure of emotional distress with acceptable test-retest reliability among patients with a vestibular disorder [50],[51].

Clinical and socio-demographic characteristics were assessed with a questionnaire that had been developed for clinical use at the Interdisciplinary Centre for Vertigo and Balance Disorders at University Hospital Zurich.

Descriptive statistics were calculated as means with standard deviations, or as percentages. The relationships between SIS, DHI, the three DHI subscales, HADS score, and clinical characteristics were analysed using Pearson correlation coefficients. Further correlation analyses were conducted to evaluate the relationships between SIS and additional variables (e.g., age). To compensate for multiple testing, a Bonferroni correction was applied to the results of the 24 correlations, the threshold for significance thereby adjusted to 0.05/24 = 0.002. Inter-group comparisons of SIS for different nominal variables like gender were performed using Student’s t-tests and one-way analysis of variance (ANOVA). Hierarchical regression analyses were conducted to measure the influence of different variables on SIS variance, with SIS as the dependent variable. All variables identified as statistically significant on bivariate analysis were entered into the models. The DHI total score model was constructed by introducing (variable insertion method: Enter) the following variables step by step into the model: a) DHI; b) DHI and continuous dizziness vs. transient attacks (code: 0 continuous, 1 transient attacks) and c) DHI, continuous dizziness vs. transient attacks and HADS score. The DHI subscale model was similarly constructed by introducing the following variables step by step into the model: a) DHI subscales; b) DHI subscales and continuous dizziness vs. transient attacks; and c) DHI subscales, continuous dizziness vs. transient attacks, and the HADS score. The Durbin-Watson test for hierarchical regression (1.76 and 1.84) suggested independent errors [35]. The average variance inflation factor was not substantially greater than 1.0 (between 1.1 and 1.6), suggesting that the regression was not biased by multicollinearity [36],[37]. To compare patients who marked their dizziness inside their self-circle against those who marked it outside, t-tests were performed. The threshold for statistical significance was set as p < 0.05, unless otherwise specified (e.g., adjusting for multiple testing). All calculations were performed using the statistical software-package SPSS (version 22).

Two hundred and three patients seen at the Interdisciplinary Centre for Vertigo and Balance Disorders at University Hospital Zurich provided their written consent and agreed to complete both the questionnaires and the PRISM tool. However, of these 203 patients, 26 neglected to complete the PRISM, leaving 177 subjects for analysis. These 177 did not differ from the 26 excluded for a missing PRISM in terms of mean DHI total score, HADS score, age, duration of symptoms, or gender distribution (all p > 0.1, data not shown). The mean age of participants was 44.4 years, 86 (48.6%) were male, and the mean SIS was 27.5 mm. The mean duration of dizziness was 161.9 weeks. Testing for skewness (3.5) indicated a non-symmetrical distribution with data skewed to the right (median: 55.5 weeks, IQR: 24–173). Further socio-demographic and clinical characteristics are shown in Table 1.

SIS was inversely correlated with the DHI total score, DHI sub-scores, and the HADS score (all p < 0.001). These correlations remained significant even after Bonferroni correction. SIS did not correlate with the duration of dizziness symptoms (Spearman’s rho = −0.05; p = 0.54) or age (r = 0.03; p = 0.70) (Table 2). Vertigo characteristics of dizziness and SIS exhibited significant associations, but did not survive Bonferroni correction.

There were no differences in mean SIS between female and male subjects, or between those living with someone and those living alone. There also were no associations between SIS and either employment status or level of education (Table 3).

Forty-five of the 177 patients placed their dizziness mark inside their self-circle (low SIS, high suffering). These patients reported significantly more severe dizziness than patients who placed the mark outside their self-circle (DHI total: t = 5.19; p < 0.001, mean difference = 19.62; SE difference = 3.78). DHIF, DHIP, and DHIE also were significantly higher in patients who marked their dizziness inside their self-circle versus those who marked it outside (mean difference from 4.1 to 8.6, all p ≤ 0.001). In addition, these 45 patients suffered from significantly greater emotional distress (HADS: t = 4.24; p < 0.001, mean difference = 7.57; SE difference = 1.36).

When total DHI score was introduced alone into the hierarchical regression model, it explained roughly thirty-four percent of the variance in SIS. When, in the second and third models, whether dizziness was continuous or transiently episodic and the HADS score were included, each variable explained an additional two percent of SIS variance. In the third model, severity of dizziness and SIS variance exhibited the strongest association (β = −0.45; p < 0.001). Continuous dizziness vs. transient attacks and the HADS rating exhibited smaller but still significant associations as well (Table 4).

On DHI sub-score analysis, during which DHI sub-scores DHIF, DHIP, and DHIE were entered into the regression model, roughly thirty-nine percent of SIS variance was explained by these variables, with DHIE as the only significant variable (p ≤ 0.01). The second and third models, again introducing continuous versus transient dizziness and the HADS rating, also contributed significantly to SIS variance (Table 5), with small contributions from both variables.