Patient selection was done between March 2020 and January 2021, composed of 120 patients who completed the three evaluations, where 99 (82.5%) were females, 21 (17.5%) males and the average age was 37.7 years. 39% had intermittent allergic rhinitis and 48% moderate asthma; more than half of the latter had access to adequate pathology control (Table 1).

The scale´s original author provided an official translation to Spanish by official translators, native Spanish speakers and bilingual. The researchers carried out a semantic evaluation of the Spanish version and after some changes, an official translator made the counter-translation from Spanish to Portuguese with review and approval by the original author.

The CARAT10 scale was self-administered in Spanish to ten patients (Fig. 3). Afterwards, an evaluation of each item was carried out as per proposed guidelines by the FACIT organisation for trans-cultural adaptations22. The average time per test were 8.7 min (DS 3.16) and all questions were verified to have an answer. Patients found the language to be clear, easy to understand and they were included in the total sample.

Scale for asthma and allergic rhinitis control CARAT-self-administered in Spanish.

Based on recommendations by the Consensus-Based Standards for the Selection of Health Measurements (COSMIN), the validation of a scale was performed with a reliability evaluation, validity and sensitivity to change of the instrument, as well as establishing the values for the minimal clinically important difference.

The reliability was assessed in two ways: internal consistency and measurement error.

Internal consistency evaluated by Cronbach´s alpha was of 0.82 for allergic rhinitis during T1(IC 95% 0.77–0.88), asthma of 0.82 (IC 95% 0.77–0.87) and global of 0.83 (IC 95% 0.79–0.88). There were no statistically significant differences between T2 and T3, obtaining a good level among visit sequences. This was similar to that yielded in the original validation, which was 0.85, and the German validation, which was 0.877,11. The Spanish CARAT (total and domains) showed satisfactory internal consistency, which was comparable to Portuguese and German validation. A Spearman´s alpha correlation, Kendall and polychoric were carried out without significant differences. McDonald´s omega was done in T1 0.88 (IC 95% 0.87–0.94), which was considered a good level without variation of the other visit sequences.

Reliability test–retest was determined by means of the intraclass correlation coefficient and the Bland Altman graphs whose concordance limits were 10.32 and 10.76 in 46 patients with a critical difference of 10.54 between T1 and T3.

Additionally, a Pearson correlation test was applied with a result of 0.6 (IC 95% 0.32–0.7) between T1 and T2 on 35 patients; 0.73 (IC 95% 0.55–0.84) on 46 patients between T2 and T3 obtaining an accurate measurement with moderate correlation after 4 weeks with ample confidence intervals.

The standard error of measurement (SEM) was 3.5 with p < 0.005 calculated on 28 participants and it was statistically significant similar to that found in the Van der Leeuw study, which was 2.8511. The minimal detectable difference was 9.6, which was high for the scale´s total score.

Validity was obtained with factorial analysis, longitudinal validity, discriminant validity and concurrent criterion validity. The number of assessed patients must be taken into account and the scales compared with CARAT for results analysis. Not all of the 120 patients were involved in the total analysis, which leads to more ample confidence intervals.

Evaluation of asthma was carried out comparing it with the ACQ5. This scale presents two different considerations from CARAT10, which may affect correlation and validation. The first one are three levels of asthma classification: adequate control, partially controlled and inadequate control, while the CARAT10 only has two: controlled and not controlled. The second is the ACQ5’s evaluation is transverse and the patient’s condition is asked at the time of evaluation; CARAT10 has a longitudinal assessment estimating the clinic only in the previous 4 weeks. However, ACQ5 is the scale validate in Spanish are most widely used by pulmonologists.

CARAT10 was compared with VAS regarding allergic rhinitis having 60/100 as cut-point, where above 60 is not controlled and below 60 is controlled. The VAS, as previously explained with ACQ5 (evaluates symptoms at the time of consultation) presents differences in the temporality of the evaluation. Structural equation analysis were used for confirmatory factorial analysis, confirming the structure of two factors as in the original article7

Correlation coefficients were used for longitudinal validation between score differences and repeated measurements of CARAT10, ACQ5 and VAS during T1-T2, T2-T3 and T1-T3 in 120 patients. The coefficients were not high; the temporality evaluation between scales was different, which may explain the outcome (Table 2).

It must be taken into account for discriminant validity that CARAT10 assesses the upper airway (rhinitis), lower airway (asthma) and unified airway (rhinitis and asthma). Results are presented categorised in three groups compared with cut points for VAS and ACQ5. Comparisons were performed on all evaluated patients during T1 using t-tests with 0.05 significance levels and two-tail hypothesis trials. ROC curves (operative characteristics of the trial) were carried out on the cut points (Fig 4) and assessed with the Wilcoxon rank test, which was statistically significant. Those results are in Table 3; cut points can be visualised on the CARAT10 in Spanish with its respective sensitivity, specificity and positive and negative predictive values.

a VAS allergic rhinitis + Asthma vs. CARAT10. b VAS (60) allergic rhinitis vs. CARAT10 allergic rhinitis. c VAS (30) allergic rhinitis vs. CARAT10 allergic rhinitis. d VAS (60) Asthma vs. CARAT10 asthma. e ACQ5 vs. CARAT10 asthma. f Graphic curve ROC. ACQ5 vs. CARAT10 asthma.

The VAS-global analysis (asthma+allergic rhinitis) vs. CARTA10-global yielded a cut-point of 18. In this case, a value of CARAT10 < 18 meant that both pathologies were not controlled; a negative predictive value (NPV) of 91% can be interpreted that out of 100 patients evaluated with CARAT10 with a result ≥ 18 (which would be a patient with both conditions controlled), 91 would be controlled upon comparing VAS < 60 (Table 3). Analysing the positive predictive value (PPV) of 50%, among 100 patients with a CARAT10-global < 18 (both conditions not controlled), in reality 50 would not have both of them controlled compared with the VAS-global > 60.

In the upper airway category (CARAT10-rhinitis) vs. VAS-rhinitis the analysis was with two cut points different in the VAS: 30 and 60. These were 8 and 7, respectively, and upon raising it one point (from 7 to 8), it presented a better performance with a positive predictive value of 90%. With a cut-point < 7 the positive predictive value diminishes to 50%, confirming that the best cut-point was 8.

The lower airway evaluation comprised two analyses, comparing CARAT10-asthma vs. VAS-asthma and CARAT10-asthma vs. ACQ5. The CARAT10-asthma vs. VAS-asthma evaluation had a cut-point in the VAS > 60, understood as an uncontrolled condition. In the CARAT10-asthma corresponds to a cut-point < 13 and a high negative predictive value of 90%.

The comparison of CARAT10-asthma vs. ACQ5 was with two analyses of the three categories of the latter. The first one compared a cut-point ACQ5 > 1.5 meaning patients with an inadequate control of asthma, equivalent to a cut-point of 13 in the CARAT10-asthma and a negative predictive value of 93%. In this case, CARAT10-asthma < 13 means an inadequate control and values ≥ 13 are understood as partial or total control.

The comparison of a cut-point ACQ5 > 0.75 represents patients with partial or inadequate asthma control. An equivalent point of 16 was found in CARAT10 with a negative predictive value of 90%. In this case, CARAT10-asthma values < 16 are interpreted as an inadequate or partial control and values ≥ 16 mean an adequate control. It is more useful to clearly differentiate patients with an adequate asthma control so a CARAT10-asthma cut-point of 16 in clinical practice is suggested.

In the concurrent criterion validity, the correlation between CARAT10, ACQ5 and VAS was measured in T1 and all 120 patients, applying the scales simultaneously (Table 4). Correlation of CARAT10-asthma and ACQ5 was acceptable and allergic rhinitis-VAS was low. The correlation between CARAT10-global with VAS-rhinitis+asthma was 0.63, existing a concurrent criterion validity with similar results in T2 and T3.

The minimal clinically important difference was determined using the distribution anchor method21 (−1, 0, +1: no difference; −3, −2, 2, 3: minimal difference; −5, −4, 4, 5: moderate difference; −7, −6, 6, 7: large difference.) The average minimal clinically important difference with CARAT10 was 3.25 (SD 3.77) similar to the German validation of 3.511.

One of the strengths of this study was its design based on the COSMIN guides for validation of scales. The calculated sample size was 120 patients, had a diagnosis of asthma and allergic rhinitis, all of whom completed the study. The results of validation are only applicable to patients diagnosed with both pathologies.

This is the first CARAT validation study that established the cutoff points to determine if rhinitis, asthma and both pathologies are controlled. It is a useful measurement tool in daily practice in patients with asthma and allergic rhinitis.

CARAT Spanish scale validation has been carried out with an adequate and rigorous clinical an statistical design, following COSMIN guides, the finding found have internal validity and can be extrapolated to the Spanish speaking population.

In conclusion, CARAT10 scale in Spanish presents a high internal consistency, good discriminant and concurrent validity. A standardised measurement method allows simultaneous evaluation of asthma control and allergic rhinitis.

Important results for clinical practice were established in this study such as the minimal clinically significant difference and cut points for allergic rhinitis control, asthma and global.

The minimal clinically important difference was 3.25 (SD 3.77) similar to that with the German validation of 3.511. This means a change in CARAT’s global score > 3.25 is the minimal difference detected by the said scale, which a patient considers as important in the control of allergic rhinitis and asthma.

The different cut-point in order to establish the control of allergic rhinitis, asthma and global were determined for the CARAT10 Spanish. The point for a controlled disease was CARAT10-rhinitis ≥ 8, sensitivity of 80% and specificity of 72%; CARAT10-asthma > 16, with sensitivity and specificity of 76%. and CARAT-global ≥18, with sensitivity and specificity of 76%.