Study design and patient population
This was an observational, prospective, multicentre study conducted in the dermatology and allergy departments of several Spanish hospitals under conditions of usual clinical practice. Data collection was performed between October 2013 and May 2014. Patients were included consecutively in the study if they were ≥ 18 years of age with a diagnosis of CSU, defined as the daily or almost daily occurrence of generalized hives or angioedema for at least 6 weeks prior to inclusion. Patients were excluded from the study if they had acute urticaria, urticaria vasculitis or other forms of urticaria not associated with the chronic form of the disease, any form of inducible CU that was not associated with CSU, angioedema without the presence of hives, pruritus related to dermatitis or other skin disease, any systemic disease or other conditions which might hinder data collection or interpretation. All patients included in the study gave their written informed consent to participate. Patients were followed up for a period of 6 weeks from inclusion in the study and were managed according to the criteria of the attending physician following their usual practice.
Patient-reported outcomes measures (PROM)
The once-daily UAS measures urticaria activity in terms of itch severity and number of hives over the past 24 h. Response options for itch severity are 0 = None, 1 = Mild (present but not annoying or troublesome), 2 = Moderate (troublesome but does not interfere with normal daily activity or sleep), 3 = Intense (severe itching, which is sufficiently troublesome to interfere with normal daily activity or sleep). Response options for hives are 0 = None, 1 = Mild (< 20 hives/24 h), 2 = Moderate (20–50 hives/24 h), 3 = Intense (> 50 hives/24 h or large confluent areas of hives). Scores on these two items are summed to create a total daily UAS score (range: 0–6 points). A Spanish version of the instrument was produced for use in the present study following standard procedures of translation and cultural adaptation of patient reported outcome (PRO) measures [
13]. The version is shown in
Appendix.
Patients completed the UAS over 7 days following the first visit and in the week prior to the second visit. Patients were asked to complete the questionnaire just before going to bed in the evening. Summing daily non-missing values UAS scores over the seven-day period provides the UAS7 score, with a score range from 0 (no activity) to 42 (most intense activity).
Other PROs included in the study were the Chronic Urticaria Quality of Life scale (CU-Q
2oL) and EQ-5D-3 L. The CU-Q
2oL is a disease-specific HRQOL questionnaire for use in patients with chronic urticaria [
14]. It consists of 23 items assessing HRQOL in 6 dimensions: itching (2 questions), swelling (2 questions), impact on daily activities (6 questions), sleep problems (5 questions), limitations (3 questions) and aesthetic problems (5 questions). Questions are answered on a 5-point Likert scale and scores are transformed to a 0 to 100 scale, with higher scores indicating poorer quality of life. The questionnaire generates both an overall score and scores by dimension and has been validated in Spanish [
15].
The EQ-5D-3 L is a widely used generic questionnaire designed to assess health status in a wide range of conditions in adult populations [
16]. It consists of 5 dimensions (mobility, personal care, daily activities, pain/discomfort, and anxiety/depression) with three possible levels of response in each dimension (absence of problems, moderate problems, extreme problems) to assess self-perceived health status on the day of completion. Utility indices based on the preferences of the general population for each of the 243 states defined by the descriptive system provide a summary score of self-rated health on the 5 dimensions and are available for various countries. They provide values on a scale anchored at 0 (death) to 1 (full health) and the Spanish value set was used to calculate the utility score in the present study [
17]. Patients also assess their overall health on a vertical visual analogue scale (EQ-VAS) from 0 to 100 where 0 represents the worst imaginable health state and 100 represents the best imaginable health state. The validated Spanish version of EQ-5D was used in the present study [
18].
The Physician’s In-Clinic UAS, which provides a rating of the patient’s itching and hives (measured on a 0–6 scale), was also completed by the attending clinician, with input from the patient, at both study visits. Furthermore, a 5 item Likert scale with responses of ‘Very mild’, ‘Mild’, ‘Moderate’, ‘Severe’, and ‘Very severe’ was used to obtain the clinician and patient’s opinion on the overall severity of their CSU in both study visits. In the final visit, a categorical scale (Global Index of Change, or GIC) was completed by physicians and patients to obtain their opinion of the evolution of the CSU over the study period. The GIC consisted of 13 response options ranging from ‘Very much worse’ to ‘Very much improved’. Patient opinions on the ease of completion and relevance of the scale for assessing their urticaria were also evaluated using a 5-point Likert scale.
Additional study variables collected included age, sex, weight and height, ethnicity, educational level, presence of chronic spontaneous urticaria (CSU), presence of other types of urticaria associated with CSU, years since onset of symptoms, years from diagnosis, number of episodes in the previous year, presence of angiodema, number of exacerbations in the previous year, whether receiving treatment or not, presence of co-morbidities associated with CSU, history of other atopic diseases, exacerbation of CSU through NSAIDs, thyroid pathology, and associated autoimmune conditions.
Statistical analysis
Descriptive analysis of all variables was performed using absolute numbers and frequencies in the case of categorical variables and means or medians together with standard deviations or interquartile ranges in the case of continuous quantitative variables. The chi-square test was used to study the relationship between qualitative variables. Parametric (Student t test or ANOVA) or non-parametric (U-Mann Whitney or Kruskal Wallis) tests were used to study the relationship between continuous variables, depending on whether distribution was normal or non-normal. A statistical significance level of p < 0.05 was used in all analysis. Statistical analyses were carried out on the study population with valid data, without any type of imputation of missing data.
The feasibility and psychometric properties (reliability, validity, and responsiveness) of the UAS were analysed using a variety of approaches.
Feasibility was assessed by analysing the proportion of missing or unusable responses per item at each visit as well as the proportion of patients with at least one missing or unusable response. We also calculated floor and ceiling effects, i.e. patients scoring the minimum (0) and maximum scores for the UAS and UAS7 (6 and 42, respectively) scores at each visit. Floor and ceiling effects under 15% are usually considered acceptable [
19].
The reliability of the questionnaire was assessed by estimating internal consistency (i.e., the degree of homogeneity of the 2 items forming the scale) and by assessing test-retest reliability (i.e., the degree of stability of the score when there are no changes in health status) [
20]. The internal consistency of the scale was evaluated by calculating the Cronbach’s alpha for the overall instrument score. Test-retest reliability was assessed by calculating the intraclass correlation coefficient (ICC) for patients who, in the second study visit, declared that their urticaria symptoms had not changed from baseline using the global change item. For both Cronbach’s alpha and the ICC, a value of 0.7 or more was considered to indicate an acceptable level of reliability [
21].
Construct validity, i.e. the degree to which the instrument performs as expected in meeting a series of pre-defined hypotheses, was assessed by evaluating its convergent and discriminant validity, as well as its known groups’ validity.
Convergent and discriminant validity refer to whether the instruments shows expected patterns of correlations with instruments measuring similar (convergent) and dissimilar (discriminant) constructs. They were assessed here by analysing the correlations of UAS7 scores with scores on the EQ-5D-3 L and CU-Q2oL at baseline. We also analysed the correlations between individual items (itch and number of hives) on the UAS and dimension scores on the CU-Q2oL. A series of hypotheses were developed regarding the expected pattern of correlations between the different instruments. The UAS7 score was expected to show lower correlations (r between 0.2 and 0.5) with the EQ-5D-3 L Index and VAS than with the CU-Q2oL global score, as the latter is a condition-specific measure. Higher correlations were expected between the UAS7 score and the EQ-5D-3 L dimensions of usual activities, pain/discomfort, and anxiety/depression than with the dimensions of mobility and self-care, as the former were considered more likely to be affected by CSU. On the CU-Q2oL, we expected moderate to high correlations (r = 0.5–0.8) between the UAS7 score and all dimensions of the CU-Q2oL, but we expected particularly high correlations between the itch items on the two questionnaires.
Known groups’ validity was assessed by determining the extent to which the instrument discriminated between patients according to their self-rating on the global CSU severity scale. The comparison was performed using ANOVA and between group effect sizes, which were calculated as the difference between group means divided by the pooled standard deviation. Effect sizes were classified as small (ES of approximately 0.20), moderate (ES of 0.50) or large (0.80 or more) [
22]. In addition, scores were calculated for the UAS7 according to the clinician rating on the In-Clinic UAS at the first visit using the categorization proposed by Mathias et al. of < 4, 4, 5, 6 points [
11].
Responsiveness was investigated by analysing the magnitude of the change in the UAS7 score corresponding to different levels of patient- and clinician-reported improvement on the global urticaria rating scale. The global index of change was also used for this analysis.
Finally, the MID for the UAS7 score was estimated in several ways, using both anchor- and distribution-based approaches [
23]. For the anchor-based approach, clinician and patient ratings of the overall severity of CSU were used as the reference and a linear regression model was constructed to estimate the score change on the UAS and UAS7 corresponding to a one category shift on the scale between the two visits. Patient ratings on the GIC were also used as an anchor in an alternative model. Distribution-based MIDs were estimated based on the standard error of measurement (SEM) and standardised effect size (SES) using values of 1 SEM and 0.5 SES, as recommended [
23].