Validation of the Atopic Dermatitis Control Tool (ADCT©) using a longitudinal survey of biologic-treated patients with atopic dermatitis

Background The Atopic Dermatitis Control Tool (ADCT©) is a brief patient self-administered instrument designed and validated to assess atopic dermatitis (AD) control; six AD symptoms and impacts are evaluated over the past week, including overall severity of symptoms, days with intense episodes of itching, intensity of bother, problem with sleep, impact on daily activities, and impact on mood or emotions. This study assessed the reliability, validity, and responsiveness of the ADCT in a longitudinal context, and provided thresholds to identify meaningful within-person change. Methods Data were from a prospective, longitudinal patient survey study of real-world effectiveness of dupilumab in patients with AD. Eligible patients completed a baseline survey before starting dupilumab and were followed at Months 1, 2, 3, and 6 post-initiation as they became eligible. Results Psychometric analyses confirmed internal consistency; Cronbach’s α coefficients were consistently above the threshold of 0.70 across each follow-up; item-to-total correlations were above the threshold of r ≥ 0.50. High correlations between the ADCT and the Dermatology Life Quality Index (DLQI) and skin pain supported construct validity, while known-group validity was shown on Patient Global Assessment of Disease (PGAD) overall well-being subgroups with worse AD-related overall well-being having higher mean ADCT total scores at all time points. The ability of the ADCT to detect change was confirmed; the threshold for meaningful within-person change was estimated to be 5 points. Finally, test–retest reliability was confirmed in subgroups of patients with stable PGAD responses. Conclusions Our findings confirm that the ADCT is a valid and reliable tool for assessing AD control.

control has been described in various ways in the literature, ranging from reduced disease severity or the absence of AD flares, to the impact of AD on patients' everyday lives and well-being [25][26][27][28][29].
The Atopic Dermatitis Control Tool (ADCT©) is a new PROM designed to assess patient-perceived disease control, meeting this current measurement gap in the management of patients with AD (ADCT v1; https:// patient-questionnaires.sanofi.com/questionnaires/adctatopic-dermatitis-control-communication-tool) (Pariser D, Simpson E, Gadkari A, Bieber T, Margolis D, Brown M, Nelson L, Mahajan P, Reaney M, Guillemin I et al: Design, validation and scoring of the Atopic Dermatitis Control Tool (ADCT), unpublished). It is envisaged that the tool will also foster patient-clinician communication regarding disease control. The ADCT is a simple, brief tool that evaluates six symptoms and effects associated with AD over the past week. These include overall severity of symptoms, days with intense episodes of itching, intensity of bother, problem with sleep, impact on daily activities, and impact on mood or emotions. Each of the six ADCT items has a score range from 0 (no problem) to 4 (worst), rating the severity of each concept; the total score ranges from 0 to 24, which is the summation of the responses to all the items. An initial evaluation of the psychometric properties of the instrument in the United States has indicated that the ADCT is valid and reliable for assessing patient-perceived AD control in adults (Pariser D, Simpson  In addition, a score of ≥7 points was derived as the threshold to identify patients "not in control", based on optimal sensitivity/specificity values (Pariser D, Simpson E, Gadkari A, Bieber T, Margolis D, Brown M, Nelson L, Mahajan P, Reaney M, Guillemin I et al: Design, validation and scoring of the Atopic Dermatitis Control Tool (ADCT), unpublished). The ADCT© is protected by copyright with all rights reserved to Sanofi and its development partners.
The present study further assessed the reliability, validity, and responsiveness of the ADCT. In addition, it defines a threshold to identify meaningful within-person change. The assessments were conducted on data from the EaRly REal-WorLd PatIent EValuation for DupixEnt in Atopic Dermatitis (RELIEVE-AD) study, a prospective, longitudinal patient survey conducted in the United States that aims to evaluate the early effectiveness of dupilumab in the real-world setting. Dupilumab is approved in the United States and Japan for subcutaneous administration every 2 weeks for the treatment of patients aged ≥12 years with moderate-to-severe AD inadequately controlled with topical prescription therapies or when those therapies are not advisable [30], and in the European Union for use in adults with moderate-tosevere AD who are candidates for systemic therapy [31].

Data source
RELIEVE-AD is an ongoing observational, prospective, longitudinal survey study in adult patients with AD who were enrolled in the Dupixent MyWay™ Patient Support Program and for whom dupilumab had been recently prescribed. Eligible patients completed a baseline survey before starting dupilumab and were followed at Months 1, 2, 3, 6, 9, and 12 post-initiation as they become eligible.
Patient enrollment into the RELIEVE-AD study began in January 2018 and the final data collection is expected to be completed in February 2020. The present study included patients in the RELIEVE-AD study who, on December 6, 2018, had completed the baseline and Months 1, 2, 3, and 6 surveys. Patients were eligible for inclusion in the RELIEVE-AD study if they met the following criteria at the time of the baseline survey: aged ≥18 years can speak and read English be willing to participate in the study and provide informed consent have not previously participated in a dupilumab clinical trial have not initiated treatment with dupilumab.
The surveys collected data on patient characteristics, including socio-demographics (age, sex, race/ethnicity, marital status, level of education, insurance, employment status, level of income, geographic region), medical history (self-reported age at AD diagnosis, comorbidities), and AD treatment and experience (treatment history prior to dupilumab initiation, concomitant therapy post dupilumab initiation, self-reported adherence to treatment and reasons for discontinuation), and treatment satisfaction.
In addition, PRO data were collected using the Patient Global Assessment of Disease (PGAD), Numerical Rating Scale (NRS) for patient self-reported symptoms (skin pain, burning, and sensitivity) (scores: 0-10; higher scores indicate worse symptom severity), disease control using ADCT (eczema-related symptoms, days with intense itching, overall bothersomeness, sleep problems, daily activities, mood/emotion; total score 0-24; higher scores indicate worse disease control), health-related quality of life (HRQL) using the Dermatology Life Quality Index (DLQI: 0-30, higher scores indicating worse HRQL), and the Work Productivity and Activity Impairment-Atopic Dermatitis questionnaire (WPAI-AD; percentages: 0-100, higher percentages indicate greater impairment) for patients in employment. (Table 1).
Analyses in this study were conducted using PRO data from multiple survey timepoints to ensure the robustness of the findings.

Statistical methods
All data analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

Reliability
Assessments for reliability included internal consistency reliability and test-retest reliability. Internal consistency was assessed using Cronbach's alpha (α ≥0.7) [32]. ADCT item-to-total correlations were estimated at baseline and Months 1, 2, 3, and 6 using the Pearson correlation coefficient (PCC ≥0.5) [33]. Test-retest reliability was evaluated based on the intra-class correlation (ICC) coefficient of the ADCT total score among patients with unchanged PGAD scores across month pairs (between Months 1 and 2, Months 2 and 3, and Months 3 and 6). An ICC ≥0.70 was expected for confirming test-retest reliability [34].

Construct validity
Convergent validity of the ADCT was assessed by computing Spearman's rank-order correlation between the ADCT total score and the DLQI (total and item-level scores), skin pain, PGAD overall well-being, WPAI total work impairment (WPAI-TWI), and WPAI total activity impairment (WPAI-TAI) at baseline and Months 1, 2, 3, and 6. Given that the skin pain NRS directly measures AD-related symptoms and the DLQI includes questions on both symptoms and impacts due to skin problems, correlations between the ADCT and skin pain and DLQI were expected to be higher than the correlations between the ADCT and other measures, such as PGAD, WPAI-TWI, and WPAI-TAI. Cohen's recommended guidelines for determining small, moderate, or large effects (0.1 to < 0.3, 0.3 to < 0.5, and ≥ 0.5, respectively) were applied, and a large effect (r ≥ 0.5) was used in this study as evidence of convergent validity [35]. Divergent validity, established previously for the ADCT, was not assessed here owing to the lack of appropriate measures for use from the current study.

Known-groups validity
To confirm known-groups validity, mean ADCT total scores were compared across adjacent subgroups of patients based on PGAD responses (Excellent, Very good, Good, Fair, Poor) and categories of DLQI responses: no effect on patient life (score range: 0-1), a small effect (2-5), a moderate effect (6-10), a very large effect (11)(12)(13)(14)(15)(16)(17)(18)(19)(20), or an extremely large effect (21)(22)(23)(24)(25)(26)(27)(28)(29)(30) [36] (Table 1). Patients in a worse PGAD or DLQI band subgroup were expected to display poorer AD control (i.e., higher mean ADCT total scores, indicating more severe symptoms/ greater impact) than patients in a better PGAD or DLQI band subgroup. If the homogeneity of variance across the subgroups was rejected (p < 0.05) based on a Levene's test of equality of variance, a Mann-Whitney U test was used to compare the mean ADCT total scores between the subgroups; otherwise, t-tests were applied. Cohen's d was calculated for the standardized differences in mean ADCT total scores between subgroups and was corrected for small sample sizes when the total sample size in the two groups was below 50 [37].

Ability to detect change (responsiveness)
Responsiveness was evaluated using correlations between the change from baseline (to Months 1, 2, 3, and 6) in ADCT total score and the change from baseline in DLQI total score (Pearson product-moment). The same analysis was conducted using (Spearman's rankorder correlation) (r ≥ 0.5) for DLQI bands and PGAD scores [34].

Interpretation of change
Anchor-based and distribution-based methods were used to establish a threshold characterizing meaningful within-person change in the ADCT total score. Prior to applying the anchor-based method, the correlation coefficient between the change in the ADCT total score and the potential anchor measure was reviewed for the magnitude of association; in this study, a large effect (i.e., correlation at least 0.5) was required [38]. Once established as appropriate, univariate regression analyses accounting for repeated measures were conducted; changes in ADCT total scores from baseline was the dependent variable and changes in the anchor measure from baseline was the independent variable. The change in PGAD and change in DLQI were considered as potential anchor measures and the following anchors were selected a priori: a 1-level improvement in the PGAD; a 4-point improvement on the DLQI total score [39]; or a 1-level improvement in the DLQI band. Patients who were not likely to change were excluded: e.g., reporting PGAD = "excellent" or DLQI = "no effect" (i.e., total score of 0 or 1) at baseline. Additional analysis was conducted using the subset of patients whose AD was considered not controlled at baseline based on the ADCT total score (i.e., score > 7; Table 1 For the distribution-based approach, the half standard deviation (SD) method of the baseline ADCT scores, one-third SD, one unit of standard error of measurement (SEM), and two units of SEMs were examined. Final recommendations for thresholds characterizing meaningful within-person change and considered as a clinical important responder were made considering the anchor-and distribution-based results.

Patient population
The interim dataset from RELIEVE-AD, as of December 6, 2018, included 1010 patients who completed the baseline survey, 538 patients at Month 1, 458 patients at Month 2, 372 patients at Month 3, and 206 patients at Month 6. Patients who were eligible to receive the survey at each timepoint varied based on time elapsed since they initiated dupilumab. Accounting for the number of surveys sent out at each timepoint, the response rate ranged from between 89.8% in Month 1 to 74.4% in Month 6. The smaller sample sizes in the later followups were attributable to many patients not due for survey completion at the time of this interim data cut. Overall, patient characteristics were comparable between patients at baseline and those who had completed the follow-up surveys.
At baseline, the mean age of the patients was 47 years, and the mean age at AD diagnosis was 28 years. More than half of the population (62%) were female and the majority (74%) were White. Most patients (96%) reported experiencing flares over the previous 4 weeks at baseline. The mean skin pain NRS score was 5.9 and the mean DLQI total score was 13.4; no patients reported a DLQI score of 0 or 1. Very few patients (3.4%) reported levels of 'excellent' or 'very good' on the PGAD. The mean WPAI-TAI and WPAI-TWI were 45.8 and 40.8%, respectively. The mean ADCT total score was 15.9 at baseline.

Construct validity
The highest correlations were observed between the ADCT total score and skin pain NRS (from 0.74 to 0.83) and the DLQI total score in the follow-up surveys (from 0.80 to 0.85), supporting construct validity (Table 4). Spearman's rank-order correlations between the ADCT total score and individual DLQI items ranged between 0.37 (issues at work or studying) and 0.75 (degree of itchiness, soreness, pain or sting) at baseline to 0.12 (issues at work or studying) and 0.75 (degree of itchiness, soreness, pain or sting) at Month 6. Other item correlations ranged between 0.4 and 0.6 regardless of the timepoint.

Known-group validity
Known-group analyses indicated that PGAD subgroups with worse AD-related overall well-being had higher mean ADCT total scores (poor AD control) at all timepoints ( Table 5). The differences in mean ADCT total score between the adjacent groups were statistically significant (p < 0.01), except between 'excellent' and 'very good' at baseline and between 'fair' and 'poor' at Month 6, likely due to small sample sizes. Similarly, patients in the groups of DLQI bands with greater effect on life were associated with higher mean ADCT total scores (poor AD control) ( Table 5). All differences in mean ADCT total score between the adjacent bands were statistically significant (p < 0.05) except between the small effect and the moderate effect bands at baseline, and between the very large effect and the extremely large effect bands at Month 6. The Cohen's d effect size showed large effect across all adjacent categories except between 'excellent' and 'very good' at baseline, between the small effect and the moderate effect bands at baseline, and between the very large effect and the extremely large effect bands at Month 6.

Ability to detect change (responsiveness)
Correlational analyses confirmed the ADCT's ability to detect change (responsiveness; Table 6). Specifically, Spearman's rank-order correlation between change in ADCT total score and change in PGAD from baseline ranged from 0.54 in Month 3 to 0.60 in Month 6. Spearman's rank-order correlation between change in ADCT total score and change in DLQI bands from baseline ranged from 0.47 in Month 1 to 0.51 in Month 3. Pearson product-moment correlation between change in ADCT total score and change in DLQI total score from baseline ranged from 0.55 in Month 1 to 0.61 in Month 3. All correlation coefficients were statistically significant (p < 0.001).

Interpretation of change
Changes in PGAD, DLQI bands, and DLQI total score correlated well with change in ADCT total   Table 4 Construct validity with Spearman's rank-order correlations between ADCT total score and other patient-  Cohen's d was corrected for small sample sizes. The correction factor was (N-3)/(N-2.25) x sqrt (1-2/N) ADCT Atopic Dermatitis Control Tool, CI Confidence interval, SD Standard deviation score (r > 0.50); therefore, PGAD and DLQI were determined to be appropriate anchors. Through the anchor-based approach, 1-level improvement in PGAD or in DLQI bands, or a 4-point reduction in DLQI total score, was associated with a reduction in ADCT total score of 5.30, 5.20, or 3.90, respectively, among the overall sample, and 5.43, 5.42, or 4.03, respectively, among patients with uncontrolled AD symptoms at baseline (Table 7). Using the distribution-based approach, the half SD and one-third SD of ADCT total score at baseline were 2.72 and 1.81, respectively. The SEM of ADCT total score at baseline was 1.71 when using the overall Cronbach's α at baseline as reliability measure and 2.32 when using the ICC between Month 1 and Month 2 as reliability measure. Consequently, 2 units of SEM were 3.42 and 4.64, respectively.

Discussion
Practice guidelines recommend that during a clinical evaluation, clinicians inquire about a patient's itch, sleep, and impact on daily activity due to their AD [15]. However, no single PROM is currently available to holistically evaluate these concepts within a single tool. The ADCT was previously developed and validated to assess AD control, with the standards recommended in the PRO Guidance by the US Food and Drug Administration [40]. Not only does this PROM evaluate AD control in a comprehensive and standardized approach, but, at the same time, it can easily be completed at home or during a clinical encounter given its brevity and ability to be self-administered via paper, online, or handheld device. The ADCT is brief, straightforward, and easily scored and interpreted, providing an immediate metric to patient self-measure of their disease control, and is thus very well adapted to clinical practice. It is anticipated that the ADCT will also facilitate meaningful patient-clinician dialogue about disease control, enhancing clinical monitoring and informing treatment decisions.
The measurement properties of the ADCT based on initial evaluations have been previously published (Pariser D, Simpson E, Gadkari A, Bieber T, Margolis D, Brown M, Nelson L, Mahajan P, Reaney M, Guillemin I et al: Design, validation and scoring of the Atopic Dermatitis Control Tool (ADCT), unpublished); in the present study, we have further evaluated this novel PROM based on data from RELIEVE-AD, a real-world, prospective, longitudinal patient survey. Cross-sectional properties previously defined were confirmed within this longitudinal context. Internal consistency was met, and moreover, ICCs computed using subgroups of patients with stable PGAD responses supported the test-retest reliability of the ADCT total score. High correlations between the ADCT and DLQI and skin pain NRS supported convergent validity, while known-groups validity was shown on PGAD subgroups with worse AD-related overall wellbeing and in the groups of DLQI bands with greater effect on life having higher mean ADCT total scores at all timepoints. Separately, versus the DLQI, the ADCT was not strongly correlated with the item "impact at work or studying" but was strongly correlated with   ***p < 0.001; a Patients who reported excellent PGAD at baseline were excluded from this analysis due to lack of variability; b No patients had 0 or 1 DLQI scores at baseline and therefore no patients were excluded from this analysis ADCT Atopic Dermatitis Control Tool "impact on social or leisure activities" at each of the timepoints. Analyses revealed a very strong correlation with the item "degree of itchiness, soreness, pain or sting". The total scores were as well strongly correlated. From these findings, it appears that self-perception of disease control is not strongly associated with AD impact at work or studying but it is very strongly with HRQL. The ability of the ADCT to detect change was confirmed for use in real-world settings; the threshold for meaningful within-person change was estimated to be 5 points. Establishing this threshold allows the clinician to assess clinically meaningful changes in AD control over time based on repeated administration of the ADCT. As previously established, a total score of ≥7 on the ADCT allows a cross-sectional assessment of lack of AD control at a given timepoint (Pariser D, Simpson E, Gadkari A, Bieber T, Margolis D, Brown M, Nelson L, Mahajan P, Reaney M, Guillemin I et al: Design, validation and scoring of the Atopic Dermatitis Control Tool (ADCT), unpublished). The meaningful within-person change threshold of 5 points compliments this by allowing a longitudinal assessment of improvement in disease control of a patient over time. Finally, good stability of the ADCT in providing reliable data over time was observed through test-retest scores against subgroups of patients with stable PGAD responses.
In consideration of our positive findings on the validity and reliability of the ADCT, a few study limitations are to be noted. First, participant diagnosis of AD relied only on self-report (i.e., not confirmed by a clinician). However, all included patients were prescribed dupilumab, which was approved only for AD when patients were enrolled in the study. Regarding the sample size, the RELIEVE-AD study is ongoing, and the full dataset is still maturing; therefore, there was a reduction of patient numbers across follow-up periods that was mainly due to the number of patients who were eligible for survey completion at those timepoints by the December 6, 2018 data cut.
While our results confirmed the psychometric properties of ADCT using data from an ongoing observational, prospective, longitudinal patient survey study, additional studies that are currently ongoing will also be able to confirm the present findings and potentially address some of the limitations. For example, one study is currently ongoing including patients and physicians to determine how ADCT compares with clinical scales (e.g., Eczema Area and Severity Index, physician assessment of control). We expect that data from such a study will also be able to confirm our findings within a population of patients with physician-confirmed diagnosis of AD.
Despite the limitations, the real-world survey data used in the present study have added further evidence to initial evaluations that the ADCT is a valid and reliable tool for assessing patient-perceived AD control and may provide a useful patient-clinician communication tool on disease control in clinical and non-clinical settings.

Conclusion
Our findings confirm that the ADCT is a valid and reliable tool for assessing AD control in real-world, longitudinal settings. In addition, ongoing studies will be able to further evaluate the present findings and potentially address some of the limitations noted.